Named after the hundred-eyed watchman of Greek myth, Argus watches the education landscape: spotting new opportunities, pressure-testing the ventures we're building, and tracing every read back to the real-world signals behind it.
The evidence library: the raw signals the pipeline is watching across the education ecosystem. Every idea is built from these.
A Center on Reinventing Public Education report outlines ways to successfully transform school systems for improvement.
I've always felt that AI tutoring as we see it now is going in the wrong direction. I made Knowable as a way to see if it's possible to have a real AI tutor. Because of hardware limitations it can only be used on Macbooks 2023+. Let me know your thoughts! Comments URL: https://news.ycombinator.com/item?id=49204343 Points: 8 # Comments: 8
arXiv:2606.14695v2 Announce Type: replace-cross Abstract: Language Models (LMs) have shown remarkable potential as role-playing chatbots, delivering consistent, stylized interactions when given a specification of a character or user persona. However, applying these capabilities to real-world applications (e.g., ecosystems with numerous NPCs interacting simultaneously) exposes a critical inefficiency due to the excessive computational cost. In this paper, we question the necessity of dedicating a full, generalist model to a single persona, hypothesizing that a specific character identity relies on only a fraction of the model's total capacity. We observe that naively pruning LMs often severely degrades the role-playing performance for a specific persona; it does not distinguish between redundant knowledge and essential character traits. We propose Persona-Pruner, a framework that sculpts a lightweight role-playing model by isolating persona-specific sub-networks from a single descriptio
arXiv:2605.20247v2 Announce Type: replace-cross Abstract: Catastrophic forgetting remains a major obstacle to continual learning in large language models (LLMs) and vision--language models (VLMs). Although Mixture-of-Experts (MoE) architectures offer an efficient path to scaling, existing LoRA-based MoE continual learning methods still face a fundamental trade-off: they either isolate experts too aggressively, limiting knowledge transfer across tasks, or allow task-specific updates to overwrite important existing parameters, leading to severe forgetting. To address this, we propose CP-MoE, a continual learning framework built around a transient expert that captures early task-specific updates and guides their integration into stable experts. CP-MoE introduces a consistency-preserving routing bias, which uses the transient expert to estimate representation similarity with stable experts and steer routing towards more compatible expert selection, and a transient expert-guided regularisat
arXiv:2605.16411v2 Announce Type: replace-cross Abstract: Hallucination remains a fundamental challenge in vision-language models (VLMs), where autoregressive generation may produce linguistically plausible yet physically inconsistent or visually ungrounded responses due to likelihood maximization under joint probabilistic modeling. We propose a stage-wise preference optimization framework for hallucination reduction through targeted multimodal data construction. Rather than directly optimizing on generic instruction-following data, our approach progressively constructs hallucination-focused preference pairs near known failure boundaries. The framework emphasizes ambiguous spatial orientation, object relationships, OCR uncertainty, and adversarial false-premise training. Hallucinated negatives are generated through minimally perturbed yet visually inconsistent alternatives, enabling Direct Preference Optimization (DPO) to better separate grounded reasoning from plausible hallucination.
arXiv:2604.22191v2 Announce Type: replace-cross Abstract: In agentic workflows, LLMs frequently process retrieved contexts that are legally protected from further training. However, auditors currently lack a reliable way to verify if a provider has violated the terms of service by incorporating these data into post-training, especially through Reinforcement Learning (RL). While standard auditing relies on verbatim memorization and membership inference, these methods are ineffective for RL-trained models, as RL primarily influences a model's behavioral style rather than the retention of specific facts. To bridge this gap, we introduce Behavioral Canaries, a new auditing mechanism for RLFT pipelines. The framework instruments preference data by pairing document triggers with feedback that rewards a distinctive stylistic response, inducing a latent trigger-conditioned preference if such data are used in training. Empirical results show that these behavioral signals enable detection of una
arXiv:2604.16742v2 Announce Type: replace-cross Abstract: Scientists have long sought to accurately predict outcomes of real-world events before they happen. Can AI systems do so more reliably? We study this question through clinical trial outcome prediction, a high-stakes open challenge even for domain experts. We introduce CT Open, an open-access, live platform that will run four challenge every year. Anyone can submit predictions for each challenge. CT Open evaluates those submissions on trials whose outcomes were not yet public at the time of submission but were made public afterwards. Determining if a trial's outcome is public on the internet before a certain date is surprisingly difficult. Outcomes posted on official registries may lag behind by years, while the first mention may appear in obscure articles. To address this, we propose a novel, fully automated decontamination pipeline that uses iterative LLM-powered web search to identify the earliest mention of trial outcomes. We
arXiv:2604.03750v2 Announce Type: replace-cross Abstract: Reverse engineering (RE) is central to software security, particularly for cryptographic programs that handle sensitive data and are highly prone to vulnerabilities. It supports critical tasks such as vulnerability discovery and malware analysis. Despite its importance, RE remains labor-intensive and requires substantial expertise, making large language models (LLMs) a potential solution for automating the process. However, their capabilities for RE remain systematically underexplored. To address this gap, we study the cryptographic binary RE capabilities of LLMs and introduce CREBench, a benchmark comprising 432 challenges built from 48 standard cryptographic algorithms, 3 insecure crypto key usage scenarios, and 3 difficulty levels. Each challenge follows a Capture-the-Flag (CTF) RE challenge, requiring the model to analyze the underlying cryptographic logic and recover the correct input. We design an evaluation framework comp
arXiv:2604.01280v2 Announce Type: replace-cross Abstract: Knowledge-based Visual Question Answering (KB-VQA) requires Multimodal Large Language Models (MLLMs) to identify and combine fine-grained visual cues with retrieved textual evidence. However, retrieval often introduces noisy and partially relevant content, while images contain distracting visual regions, causing pretrained MLLMs to overlook the evidence that actually supports the answer. To address this, we introduce Look Twice (LoT), a training-free inference-time framework that turns the model's own internal attention into an explicit multimodal evidence-selection mechanism. LoT first leverages the model's internal attention patterns to identify query-relevant image regions and textual sentences, filters attention sinks and distracting content, and reformulates the input to explicitly highlight the selected evidence before answer generation. The method requires no parameter updates, auxiliary models, or architectural modificat
arXiv:2604.01170v2 Announce Type: replace-cross Abstract: While test-time scaling has enabled large language models to solve highly difficult tasks, state-of-the-art results come at exorbitant compute costs. These inefficiencies can be attributed to the miscalibration of post-trained language models, and the lack of calibration in popular sampling techniques. Here, we present Online Reasoning Calibration (ORCA), a framework for calibrating the sampling process that draws upon conformal prediction and test-time training. Specifically, we introduce a meta-learning procedure that updates the calibration module for each input. This allows us to provide valid confidence estimates under distributional shift, e.g. in thought patterns that occur across different stages of reasoning, or in prompt distributions between model development and deployment. ORCA not only provides theoretical guarantees on conformal risks, but also empirically shows higher efficiency and generalization across differen
arXiv:2601.03895v2 Announce Type: replace-cross Abstract: Group Relative Policy Optimization (GRPO) has emerged as a popular algorithm for reinforcement learning with large language models (LLMs). However, GRPO inherits PPO's token-level clipping while replacing token-level advantages with a single sequence-level advantage. Through a four-quadrant analysis of the (likelihood-ratio, advantage) space, we show that this combination leaves one quadrant -- negative advantage combined with an increased likelihood ratio (Q4) -- structurally unbounded, so that a few high-ratio tokens can receive very large suppressive updates that collapse entropy and narrow the reasoning boundary. To address this, we propose All-Quadrant Bounded Clipping GRPO (ABC-GRPO), which applies unconditional clipping in all four quadrants through sign-dependent boundaries. ABC-GRPO clips the likelihood ratio before multiplying by the advantage, adding a trust-region floor in Q2 and a cap in Q4 -- its negative-advantage
arXiv:2509.06861v3 Announce Type: replace-cross Abstract: Test-time scaling increases inference-time computation through longer reasoning chains and has shown strong performance gains across many domains. However, frontier models still suffer from factuality hallucinations, raising the question of whether increased computation is effective on closed-book knowledge-intensive tasks. In this work, we evaluate 14 reasoning models under different test-time scaling strategies. Our results challenge its effectiveness: increasing test-time computation does not consistently improve accuracy and often leads to more hallucinations. We find that changes in hallucination rates are largely driven by the model's willingness to answer, as longer reasoning encourages more attempts, many of which are incorrect. We also observe patterns consistent with confirmation bias, where extended reasoning reinforces early incorrect beliefs with fabricated details. Finally, we provide an information-theoretic persp
arXiv:2608.04586v2 Announce Type: replace Abstract: Multimodal large language models (MLLMs) have achieved significant success in speech-to-text translation (S2TT). However, when processing multilingual speech inputs, a single speech encoder shared across all languages suffers from the curse of multilinguality: languages at different resource levels compete for limited representation capacity, leading to strong high-resource performance but substantial degradation on low-resource speech. To address this problem and improve multilingual consistency, we propose MSRT, a novel framework built around a resource-aware Mixture of Speech Encoders (MoSE). MoSE uses an explicit language router to assign each utterance to an appropriate expert encoder. A frozen expert preserves high-resource language capabilities, while a trainable expert adapts to and specializes in medium- and low-resource languages. We further introduce a five-stage curriculum learning strategy that substantially reduces data
arXiv:2608.02616v2 Announce Type: replace Abstract: We present what is, to our knowledge, the first systematic evaluation of OpenAI's Privacy Filter (OPF), a 1.5B-parameter model that converts an autoregressive language model into a bidirectional PII detector, across 32 benchmarks spanning 14 languages and 5 domains. Our most practically actionable finding is a domain-dependent labeled-data crossover: fine-tuned XLM-RoBERTa surpasses OPF's zero-shot performance with only ~500 labeled examples on English synthetic PII (~100 on non-English Kiji), and ~1000 on synthetic medical PII. Crucially, per-class fine-tuning (17 PII entity types, a subset of OPF's 33) is less data-efficient than binary labels at small n -- at n=100, binary F1=0.634 vs. per-class 0.360. Zero-shot, OPF achieves F1=0.464 on the SPY medical benchmark and F1=0.855 on AI4Privacy, substantially outperforming Presidio and XLM-RoBERTa-large-NER. However, OPF degrades sharply outside its PII training distribution: F1=0.04--0
arXiv:2608.01347v3 Announce Type: replace Abstract: Two prompts can request the same code change and produce the same correct patch, yet cause a coding agent to perform radically different kinds and amounts of work. We study this effect in a preregistered benchmark spanning 4,644 valid runs, 24 deterministic coding tasks, seven reasoning models, and two real agent harnesses. The central finding is that prompt wording does not merely scale total effort; it changes where that effort is spent. Multiple approaches and deep thinking primarily inflate reasoning. Multiple approaches increases reasoning by 2.4x to 7.4x across all six open models and creates about three elaborated but discarded solution branches, while still yielding only one implemented solution and no success gain. Maximum certainty activates a different pathway: repeated verification propagates into extra test runs, tool calls, turns, latency, and context growth. Runs with high redundant verification cost 18x the clean-run m
arXiv:2606.21155v2 Announce Type: replace Abstract: Attorneys, judges, and pro se filers increasingly use AI to draft legal documents, yet these tools frequently fabricate citations. Despite predictions that newer models would hallucinate less or that court sanctions would deter negligent filers, we found over 1,000 filings containing fabricated citations---with this number growing year-over-year. This study evaluates whether AI-based systems can mitigate these errors by automatically detecting hallucinations. We propose a taxonomy of legal citation hallucinations grounded in actual court filings and introduce a dataset of 1,300 brief excerpts containing injected errors. Benchmarking five models in agentic and non-agentic settings as well as Claude Code reveals that while the latest iterations perform better---GPT-5 achieves 84.4% recall and a 55.0% F1 score in an agentic framework---all models struggle with subtle error categories. Agentic verification remains resource-intensive, with
arXiv:2606.13227v2 Announce Type: replace Abstract: Post-training methods such as supervised fine-tuning (SFT) and preference optimization typically align language models toward a single global assistant behavior. While effective for improving average helpfulness, this can suppress the natural variation of human responses across languages, tasks, and dialogue settings. We study this problem as conditional human-distribution alignment: models should match the human response distribution appropriate to the current interaction context, rather than a universal response style. We introduce PolyAlign, a distribution-aware alignment framework that organizes bilingual interaction data into bucket-specific human reference distributions defined by language, interaction track, response family, and length. PolyAlign combines Bucket-Aware SFT, which balances optimization across heterogeneous buckets, with Human-Distribution Preference Optimization (HDPO), which regularizes preference learning using
arXiv:2606.10921v2 Announce Type: replace Abstract: Long-document question answering (QA) requires large language models (LLMs) to reason over evidence scattered across lengthy documents, where answers often depend on event order, section-level context, and cross-part evidence connections. Although retrieval-augmented generation (RAG) reduces the input context by retrieving relevant evidence, existing structured RAG methods still face three limitations: costly query-agnostic knowledge organization, insufficient use of original document structure, and no reuse of historical reasoning experience. To address these limitations, we propose DocTrace, a multi-agent RAG framework for long-document QA that supports query-triggered knowledge organization, document-structure-aware and experience-guided reasoning. DocTrace preserves document hierarchy with a lightweight document structural tree index, constructs agent-shared hypergraph-structured working memory on demand during reasoning, and stor
arXiv:2606.02776v4 Announce Type: replace Abstract: When large language models (LLMs) are used in high-stakes scenarios, such as legal, medical and financial advice, even a single conversation history is enough to drive differences in outcomes between users. Prior work has demonstrated that this results in outcome disparities between sociodemographic groups, with some groups receiving more advantageous outcomes than others. In this work, we demonstrate that LLMs actually struggle to infer user sociodemographics from a single conversation history and that although there are disparities between sociodemographic groups, they are minimal in magnitude. To investigate what is the main driver of disparities between users, we compare user sociodemographics to a range of (psycho)linguistic features of conversations, including conversation topic, emotions, and readability. We find that conversation topics are most predictive of LLM-generated advice within a conversational context, which, to some
arXiv:2605.12519v2 Announce Type: replace Abstract: Training language models to produce both correct answers and sound reasoning remains an open challenge. Reinforcement learning with verifiable rewards typically optimizes only final outcomes, which can improve task accuracy at the expense of reasoning quality, producing inaccurate, incomplete, or inconsistent traces. We propose verifiable process supervision (VPS), a post-training framework that jointly optimizes prediction accuracy and reasoning quality by supervising structured intermediate claims. We first apply supervised fine-tuning to induce a structured reasoning format, enabling deterministic extraction and verification of intermediate claims for process-level rewards. To address the heterogeneous difficulty of reasoning subtasks, we introduce adaptive weighting that prioritizes components with the largest remaining errors, creating an implicit curriculum. We evaluate VPS on chess as a controlled testbed where reasoning steps
arXiv:2605.05066v2 Announce Type: replace Abstract: We identify and prove a fundamental trade-off governing long-sequence models: no model can simultaneously achieve (i) per-step computation independent of sequence length (Efficiency), (ii) state size independent of sequence length (Compactness), and (iii) the ability to recall a number of historical facts proportional to sequence length (Recall). We formalize this trade-off within an Online Sequence Processor abstraction that unifies Transformers, state space models, linear recurrent networks, and their hybrids. Using the Data Processing Inequality and Fano's Inequality, we prove that any model satisfying Efficiency and Compactness can recall at most O(poly(d)/log V) key-value pairs from a sequence of arbitrary length, where d is the model dimension and V is the vocabulary size. We classify 52 architectures published before March 2026 into the triangle, showing that each achieves at most two of the three properties and that hybrid arc
arXiv:2604.01707v3 Announce Type: replace Abstract: Memory emerges as the core module in the large language model (LLM)-based agents for long-horizon complex tasks (e.g., multi-turn dialogue, game playing, scientific discovery), where memory can enable knowledge accumulation, iterative reasoning and self-evolution. A number of memory methods have been proposed in the literature. However, these methods have not been systematically and comprehensively compared under the same experimental settings. In this paper, we first summarize a unified framework that covers existing representative agent memory methods from a high-level perspective. We then extensively compare representative agent memory methods on two long-term conversational benchmarks and an agentic memory benchmark, and examine the effectiveness of representative methods, providing a thorough analysis of those methods. As a byproduct of our experimental analysis, we also design a new memory method by exploiting modules in the exi
arXiv:2602.14265v3 Announce Type: replace Abstract: Inference-Time-Compute (ITC) methods like Best-of-$n$ and Tree-of-Thoughts are meant to produce output candidates that are both high-quality and diverse, but their use of high-temperature sampling often fails to achieve meaningful output diversity. Moreover, existing ITC methods offer limited control over $\textit{how}$ to perform reasoning, which in turn limits their interpretability. We present $\textbf{STATe-of-Thoughts}$ (STATe), an interpretable ITC method that $\textit{searches}$ over high-level reasoning patterns. STATe branches over discrete and interpretable textual interventions rather than over token-level samples: a $\textit{controller}$ selects actions encoding high-level reasoning choices; a $\textit{generator}$ produces reasoning steps conditioned on those choices; and an $\textit{evaluator}$ scores candidates to guide search. This structured approach yields three main advantages. First, action-guided textual interventi
arXiv:2601.12369v5 Announce Type: replace Abstract: Deep Research Agents increasingly automate survey writing, yet existing benchmarks do not jointly test whether they retrieve the papers experts consider essential and organize those papers into paper-grounded taxonomies. We introduce TaxoBench, a benchmark built from 72 highly cited LLM surveys, 3,815 cited papers, and their expert-authored taxonomies. TaxoBench evaluates systems in two settings: Deep Research mode measures end-to-end retrieval and organization from a topic, while Bottom-Up mode provides the expert paper set and isolates organization. We evaluate leaf-level assignments with ARI and V-Measure and hierarchy-level structure with US-TED, US-NTED, and Sem-Path. Across 7 Deep Research Agents and 16 LLM configurations, the best agent retrieves only 20.92% of expert-cited papers, and none of 70 standard Bottom-Up runs reaches the experts' average depth of 4.86. A controlled probe shows that models which match this depth do so
arXiv:2601.05192v2 Announce Type: replace Abstract: Entity linking (mapping ambiguous mentions in text to entities in a knowledge base) is a foundational step in tasks such as knowledge graph construction, question-answering, and information extraction. Our method, LELA, is a modular coarse-to-fine approach that leverages the capabilities of large language models (LLMs), and works with different target domains, knowledge bases and LLMs, without any fine-tuning phase. Our experiments across various entity linking settings show that LELA is highly competitive with fine-tuned approaches, and substantially outperforms the non-fine-tuned ones.
arXiv:2601.04098v2 Announce Type: replace Abstract: Transformer language models systematically prefer tokens at specific input positions regardless of semantic relevance---a phenomenon known as positional bias. Prior work characterizes this bias in model behavior through performance drops in long-context tasks or in model architecture through attention-based analyses. However, it remains unmeasured how input positions actually drive predictions layer by layer. We introduce a layer conductance framework within a sliding-window design, applied to short-context next-word prediction to isolate model-internal behavior from task and context-window pressure. The resulting layer-wise positional importance profiles are stable across diverse texts and lexical scrambling, confirming they reflect model-internal structure. Characterizing how these profiles evolve across depth, we find recency bias increases monotonically while primacy bias is subtle and diminishes. We also find that this positional
arXiv:2511.17241v2 Announce Type: replace Abstract: Understanding and predicting user behavior on social media platforms is crucial for content recommendation and platform design. While existing approaches focus primarily on common actions like retweeting and liking, the prediction of rare but significant behaviors remains largely unexplored. This paper presents a hybrid methodology for social media user behavior prediction that addresses both frequent and infrequent actions across a diverse action vocabulary. We evaluate our approach on a large-scale Bluesky dataset containing 6.4 million conversation threads spanning 12 distinct user actions across 25 persona clusters. Our methodology combines four complementary approaches: (i) a lookup database system based on historical response patterns; (ii) persona-specific LightGBM models with engineered temporal and semantic features for common actions; (iii) a specialized hybrid neural architecture fusing textual and temporal representations
arXiv:2510.03997v2 Announce Type: replace Abstract: Understanding how patients perceive their physicians is essential to improving trust, communication, and satisfaction. Patients increasingly consult large language models (LLMs) to summarize physician reviews and shape provider choices, yet the national landscape of patient-perceived physician traits remains poorly characterized. We present an LLM-based pipeline that extracts ten patient-perceived physician trait scores from review text: five Big-Five-style and five patient-oriented dimensions. From one million U.S. physicians, we analyze 4.1 million reviews of 226,999 physicians. We validate the pipeline through multi-model comparison and human expert benchmarking. LLM and human-rater trait scores from reviews are consistent. Trait scores correlate strongly with review rating scores yet retain substantial independent variance. Two national-scale patterns emerge: male physicians receive higher trait scores across all traits, with the
arXiv:2509.16666v2 Announce Type: replace Abstract: Large language models (LLMs) often achieve high performance in native language identification (NLI) benchmarks by leveraging superficial contextual clues such as names, locations, and cultural stereotypes, rather than the underlying linguistic patterns indicative of native language (L1) influence. To improve robustness, previous work has instructed LLMs to disregard such clues. In this work, we demonstrate that such a strategy is unreliable and model predictions can be easily altered by misleading hints. To address this problem, we introduce an agentic NLI pipeline inspired by forensic linguistics, where specialized agents accumulate and categorize diverse linguistic evidence before an independent final overall assessment. In this final assessment, a goal-aware coordinating agent synthesizes all evidence to make the NLI prediction. On two benchmark datasets, our approach significantly enhances NLI robustness against misleading context
arXiv:2507.14022v2 Announce Type: replace Abstract: This study proposes the Cognitive Pairwise Comparison Classification Model Selection (CPC-CMS) framework for document-level sentiment analysis. The CPC, based on expert knowledge judgment, is used to calculate the weights of evaluation criteria, including accuracy, precision, recall, F1-score, Specificity, Matthews Correlation Coefficient (MCC), Cohen's Kappa (Kappa), and efficiency. Naive Bayes, Linear Support Vector Classification (LSVC), Random Forest, Logistic Regression, Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and A Lite Bidirectional Encoder Representations from Transformers (ALBERT) are chosen as classification baseline models. A weighted decision matrix consisting of classification evaluation scores with respect to criteria weights is formed to select the best classification model for a classification problem. Three open datasets of social media are used to demonstrate the feasibility of the propose
arXiv:2507.04884v2 Announce Type: replace Abstract: We consider open-retrieval conversational question answering (OR-CONVQA), an extension of question answering where system responses need to be (i) aware of dialog history and (ii) grounded in documents (or document fragments) retrieved per question. Domain-specific OR-CONVQA training datasets are crucial for real-world applications, but hard to obtain. We propose a pipeline that capitalizes on the abundance of plain text documents in organizations (e.g., product documentation) to automatically produce realistic OR-CONVQA dialogs with annotations. Similarly to real-world humanannotated OR-CONVQA datasets, we generate in-dialog question-answer pairs, self-contained (decontextualized, e.g., no referring expressions) versions of user questions, and propositions (sentences expressing prominent information from the documents) the system responses are grounded in. We show how the synthetic dialogs can be used to train efficient question rewr
arXiv:2506.18203v3 Announce Type: replace Abstract: Verifiers can improve language model capabilities by scoring and ranking responses from generated candidates. Currently, high-quality verifiers are either unscalable (e.g., humans) or limited in utility (e.g., tools like Lean). While LM judges and reward models have become broadly useful as general-purpose verifiers, a significant performance gap remains between them and oracle verifiers (verifiers with perfect accuracy). To help close this gap, we introduce Weaver, a framework for designing a strong verifier by combining multiple weak, imperfect verifiers. We find weighted ensembles of verifiers, which typically require learning from labeled data, significantly outperform unweighted combinations due to differences in verifier accuracies. To reduce dependency on labeled data, Weaver leverages weak supervision to estimate each verifier's accuracy and combines outputs into a unified score that better reflects true response quality. Howe
arXiv:2506.03259v3 Announce Type: replace Abstract: Purpose: To compare five lightweight open-weight large language models (LLMs) with a rule-based algorithm (RBA) and fine-tuned RadBERT for zero-shot labeling of chest-abdomen-pelvis (CAP) CT reports, and to examine how labeling conventions affect measured performance. Materials and Methods: In this retrospective study, 40,833 CAP CT reports from 29,540 patients examined between 2012 and 2017 were analyzed; age and sex were unavailable. Five LLMs were prompted zero-shot to assign 15 labels across three organ systems and compared with an RBA and fine-tuned RadBERT. Inter-model agreement was assessed with Cohen kappa ($\kappa$) on 12,197 held-out reports. Macro-averaged F1 was computed against 1,789 radiologist-supervised annotations, the same annotations simplified to disregard clinical actionability, and the CT-RATE dataset. Nonoverlapping bootstrapped 95% CIs indicated relevant differences. Results: MedGemma 27B and MedGemma-1.5 4B sh
arXiv:2502.14671v4 Announce Type: replace Abstract: Large Language Model (LLM) representations are known to align with brain activity during language processing, but it remains unclear what drives this alignment. We test whether explainable AI (XAI) can help answer this: using attribution methods, we quantify the contribution of each input word to an LLM's next-word predictions and use these explanations to predict fMRI data from participants listening to narratives. We find that gradient-based attribution methods robustly align with brain activity, contribute unique variance beyond acoustic and word-rate confounds, and outperform internal representations in early auditory regions. Using conductance, we extend attribution from words to individual layers, asking what each layer's attribution reveals about the model's computation and how this relates to its brain alignment. Early layers show greater word-type sensitivity and align preferentially with auditory regions, whereas the final l
arXiv:2405.15604v4 Announce Type: replace Abstract: Text generation has become more accessible than ever, and the growing interest in these systems, especially those using large language models, has spurred a surge in related publications. We provide a systematic literature review comprising 257 papers, covering the period from January 2017 to December 2025. This review categorizes text generation contributions into five main tasks: open-ended text generation, summarization, translation, paraphrasing, and question answering. For each task in our taxonomy, we review relevant characteristics and key subtasks. We assess current approaches for evaluating text generation systems, covering model-free, model-based, and human evaluation. Our investigation shows several task-specific challenges (e.g., missing datasets for multi-document summarization, lack of coherence in story generation, and difficulties in complex reasoning for question answering). We further discuss nine challenges common t
arXiv:2204.07228v2 Announce Type: replace Abstract: This paper explores the integration of human linguistic insights into multilingual text-to-speech (TTS) systems by evaluating the Featurally Underspecified Lexicon (FUL) as a theory-driven input representation. Unlike data-intensive end-to-end models, FUL offers a compact, interpretable feature set grounded in phonological principles, enabling scalable and equitable TTS development for low-resource languages. We provide a mapping from language-specific phones to FUL feature vectors via a SAMPA intermediate and incorporate these features into a modified FastSpeech architecture. Experiments were conducted to evaluate their ability to generate native, non-native, and code-mixed speech in English and Mandarin. We ran an experiment with a small dataset and one with a larger dataset, which showed that TTS with FUL features as input could produce intelligible native speech with as little as 8 hours of training data; with 100 hours of trainin
arXiv:2608.06362v1 Announce Type: cross Abstract: Deciding which of two agents is stronger means playing games until skill outweighs luck, and every game costs money, model inference, or expert time. Since the number of games needed is unknown, fixed-budget evaluations either keep paying after the result is settled or stop before the agents can be told apart, while naive optional stopping with an ordinary confidence interval invalidates the stated level. We make such an evaluation stop as soon as its evidence suffices, with the guarantee intact. The Action-Informed Value Assessment Tool (AIVAT) reduces variance in imperfect-information games through conditional mean-zero corrections, by a median $54\times$ across 15 LLM agent configurations spanning 71,439 paired Heads-Up No-Limit Hold'em (HUNL) hands, but does not say when to stop. We combine AIVAT with continuously monitored Confidence Sequences (CSs) into anytime-valid AIVAT (AV-AIVAT), whose online value model learns only from past
arXiv:2608.06352v1 Announce Type: cross Abstract: Training terminal agents requires executable and verifiable tasks that are not merely solvable, but appropriately challenging for learning. Executable validation establishes feasibility, yet does not reveal how a task behaves relative to a given solver setting. In this paper, we present CalibForge, an autonomous terminal-task synthesis system that uses verified solver behavior to revise candidate tasks through adversarial solver calibration. Multi-solver calibration targets disagreement within a heterogeneous solver pool, whereas contrastive solver calibration targets a designated strong-pass/weak-fail relation; both operationalize a solver-relative learnable zone anchored in demonstrated solvability. Using CalibForge, we construct 5,431 calibrated terminal tasks. Our ablations show that both strategies yield more effective supervision than authoring and validation alone or ordinary single-solver feedback. Models trained on the full col
arXiv:2608.06310v1 Announce Type: cross Abstract: Recent advances in reward modeling show a paradigm shift from discriminative reward models to generative reward models. However, despite their strong capabilities in response ranking, generative reward models have not realized their potential in reinforcement learning (RL). Our analysis reveals that this limitation arises from a mismatch between the comparative nature of generative reward modeling and the scalar scoring paradigm adopted by existing RL algorithms. To bridge this gap, we propose a Ranking-based Reward Construction (RRC) approach, which enables generative reward models to provide more effective RL learning signals by deriving rewards from relative preference rankings. RRC introduces two complementary strategies: self-competitive ranking, which exploits comparisons among sampled responses, and anchor-guided ranking, which enables scalable ranking-based reward construction with a small set of reference responses. Experiments
arXiv:2608.06305v1 Announce Type: cross Abstract: Retrieval-augmented generation over long documents is dominated by one design: chunk the text, embed the chunks, and surface the top-k nearest neighbours of the query. We argue that for an important class of documents -- financial statements, audit reports, regulatory returns -- this design is structurally unsound, and we make the argument measurable. On a 780-page government financial report, 86.8% of content lines are table rows, thousands of near-identical figures compete in one embedding space, and a figure inherits its unit from a header a median of 13 lines above it -- so a chunk boundary routinely separates a number from whether it is in lakh or crore, an error of two orders of magnitude. A table-aware chunker built as a steelman fixes the unit problem but leaves 27-30% of numeric chunks with no fiscal-year header at every chunk size we tried. We propose READ (Reliable Embedding-free Agentic Document-search), in which an agent re
arXiv:2608.06301v1 Announce Type: cross Abstract: As LLMs are increasingly deployed within agentic systems, their capabilities depend not only on the model weights but also on the harness: the prompts, tools, control flow, memory, and orchestration code surrounding them. This makes automated harness optimization -- the iterative and evaluation-guided improvement of a harness by an AI system -- both an important route to improving AI systems and a demanding capability for AI systems themselves. Yet the community lacks a common protocol for measuring how well frontier LLMs perform at this task. We introduce HarnessOpt-Bench, a benchmark for end-to-end harness optimization under expensive and stochastic evaluation. An optimizer, an LLM paired with a coding harness, receives a target agent's seed harness, graded evaluation feedback, and a fixed target-evaluation budget. It edits the harness and nominates a final candidate, which is scored by its normalized gain over the seed on a held-out
arXiv:2608.06167v1 Announce Type: cross Abstract: We present a schema-based framework for extracting complex, structured information from unstructured text documents using generative AI, followed by automated semantic evaluation of the extracted information against a gold standard. The schema, serving as an information model encoding domain knowledge, provides a unified, systematic, and consistent framework for extraction of hierarchical, nested information, with attributes of variable cardinality, and subsequent evaluation of the results. Information extraction from a document is performed in a single call to the model, in zero-shot mode. In the evaluation step, we introduce a path-based semantic matching algorithm to align the nested, variable-cardinality attributes in the extracted results with those in the gold standard. We use generative AI for semantic comparison of the extracted and gold standard values of an attribute, and introduce a rubric to classify the result of the compar
arXiv:2608.06123v1 Announce Type: cross Abstract: Measuring political bias in large language models (LLMs) remains challenging as it can manifest through subtle differences in framing, argumentation, and legal reasoning that are difficult to capture with a single metric. In this work, we introduce Poli-Bias, a counterfactual framework for measuring whether LLMs treat legally equivalent conflict scenarios differently depending on the countries involved. Poli-Bias compares responses to paired prompts in which country identities are systematically swapped across diverse geopolitical relationships, legal violations, and reasoning tasks. Rather than reducing bias to a single judgment, our framework decomposes response disparities into five interpretable dimensions, revealing how and where unequal treatment manifests. Across 13 contemporary LLMs spanning diverse model families and sizes, we find that country identities and user affiliations can systematically affect how equivalent actions ar
arXiv:2608.06112v1 Announce Type: cross Abstract: Hospitals are rapidly adopting artificial intelligence for triage, imaging, scheduling etc., yet most deployments remain isolated point solutions locked inside departmental silos, resulting in duplicated effort, hidden risks, and unrealized enterprise value. Despite explosive growth of AI in healthcare market and accelerating investment, an estimated 70-80% of healthcare AI pilots fail to scale, largely due to governance gaps, fragmented data, and missing integration blueprints. This research proposes a hospital-specific, compliance-first, Agentic AI architecture with multiple interoperable layers, extending existing hospital AI platform models with: (i) an Agent Orchestration Layer for multi-agent workflows across clinical, operational, and financial domains, (ii) a Compliance and Policy Layer that centralizes policy-as-code for HIPAA, GDPR, the EU AI Act, DISHA Act, India's DPDP Act, and ISO/IEC security and safety standards, and (iii
arXiv:2608.06110v1 Announce Type: cross Abstract: This paper presents ECHO (Enhanced Care \& Health Observer), a locally-deployable conversational health assistant for long-term chronic care management. ECHO integrates three complementary software modules developed under shared supervision as a unified system. The core module is an agentic chatbot built on a ReAct loop orchestrated via LangGraph, equipped with 17 clinical tools and a temporal knowledge graph for persistent cross-session memory; it achieves a 94.9\% tool-execution pass rate across a 59-scenario benchmark with GPT-5 Mini. A two-stage hybrid safety layer intercepts all incoming queries: a rule-based layer handles explicit crisis signals and jailbreak attempts in under 1ms, while a signed graph neural network (GNN) with APPNP-style propagation classifies boundary cases by clinical intent, achieving 88.8\% accuracy and 90.6\% unsafe recall on a 2,537-query annotated Turkish health dataset while outperforming zero-shot LLM b
arXiv:2608.06041v1 Announce Type: cross Abstract: Large language models (LLMs) have been shown to exhibit strong Python preferences when generating project-level code, but there is currently no systematic way to measure this behaviour across new models. To bridge this gap, we introduce LangChoiceBench, a project-level code-generation benchmark for measuring Python preference, recommendation-implementation consistency, and language diversity. LangChoiceBench covers 28 projects across seven software areas where Python is often a poor default. We evaluate 25 diverse LLMs and find that Python remains heavily over-selected, recommendation-implementation consistency is low, and smaller open-weight models generally show stronger Python preference and lower language diversity. We further analyse 9,826 reasoning traces and find that most Python choices are automatic or driven primarily by ease, rather than explicit consideration of project requirements. In a smaller but important set of cases,
arXiv:2608.05891v1 Announce Type: cross Abstract: Mobile GUI agents can operate apps through pixel perception and touch actions, making them a promising interface for collecting and improving long-horizon mobile interaction policies. However, real trajectories are difficult to obtain for sensitive apps and privacy-critical operations. At the same time, existing simulated environments are costly to scale up, and GUI world models still suffer from unstable generation, limited modality coverage, and inconsistent action-transition logic. To address these limitations, we propose AppDeltaWorld, a transition-grounded delta code world model that predicts the next GUI as a reachable code update rather than as an unconstrained image or text description. AppDeltaWorld retrieves app-specific Level-1 HTML references under an action-transition constraint, generates Level-2 executable HTML conditioned on the current screen, action, predicted next-screen text, and retrieved structure, and inserts gene
arXiv:2608.05884v1 Announce Type: cross Abstract: Existing guidance identifies excessive agency, excessive permission, weak task-bound authorization, and inadequate agent controls as important risks. Control frameworks also describe capabilities for constraining, authorizing, observing, validating, and responding to agent activity. Yet security programs still need a way to manage persistent deployed instances that span components and outlive any one event. We propose the agentic posture vulnerability (APV) as a task-conditioned vulnerability-management abstraction: a durable record for a composed agent-control exposure. One posture may produce different runtime manifestations across tasks; APV links those manifestations to the invariant posture and remains open until authority is narrowed, a missing control is added, risk is accepted, or closure is verified. APV is not proposed as a new root-cause class of risk; it operationalizes existing excessive-agency, authorization, and control-c
arXiv:2608.05876v1 Announce Type: cross Abstract: User requests serve as research specifications for deep research agents, shaping what evidence to seek and how to synthesize it. In personalized deep research, these specifications must additionally reflect user goals, constraints, preferences, and evaluation criteria. User context can be incorporated either within the deep research pipeline or into the research specification provided as its input. We focus on the latter, refining the user request into a personalized research specification before passing it to an unchanged deep research agent. This requires resolving three coupled decisions: which framing factors are relevant, whether the available user context sufficiently supports them, and whether to retrieve user memory, ask the user, or stop and refine the query. For training, G-STEER organizes framing factors as elicitation targets in an Intent Elicitation Graph that captures their dependencies. It learns a clarification policy fr
arXiv:2608.05810v1 Announce Type: cross Abstract: Self-evolving agents accumulate capability by distilling reusable skills from their execution trajectories, but we find this process is not monotonic: past a critical pool size, newly added skills degrade performance instead of improving it. We formalize this capability-contamination phase transition and trace it to a structural cause: once a defective skill enters the decision context, it becomes reference material for distilling later skills, forming cross-round contamination chains. We further show the contamination is structurally irreversible: removing a source skill after the fact cannot erase the flawed reasoning its descendants have already inherited, so post-hoc rollback recovers only a small fraction of the lost performance. This makes skill admission a pre-commit necessity rather than a post-hoc fix, and motivates Verifier-as-Gatekeeper (VaG): a progressive trust hierarchy whose three heterogeneous critics - structural validi