Daily Papers

Generative AI systems (GenAI) are increasingly used by students as learning companions, yet little is known about how they regulate these tools in open-ended learning settings, where the goal is to improve understanding for later independent performance. This study examined Grade-9 students' use of a general-purpose GenAI during mathematical-modeling practice through the lens of epistemic proactivity (EP)---the extent to which learners retain responsibility for regulating and advancing their own knowledge-building. We operationalized EP through two complementary behavioral dimensions: self-regulated learning (SRL) functions and help-seeking (HS) content. We asked whether static summaries of these behaviors or their within-session development were more informative about subsequent unaided performance. A total of 112 students interacted with a web-based LLM tutor, producing 880 substantive student turns; 95 completed both the unaided pre- and post-tests. Interactions were strongly request-dominated, with comparatively little explicit planning, monitoring, or evaluation, and students' stated HS intentions were only weakly reflected in their enacted behavior. Static summaries of AI use did not account for additional post-test variance beyond pre-test performance. In contrast, among the 89 students with HS computable trajectories, within-session HS development was informative: students whose requests became relatively more understanding-oriented---toward conceptual and procedural assistance rather than answer-seeking or verification---showed higher subsequent unaided performance. These findings suggest that AI-supported learning cannot be understood from aggregate interaction characteristics alone; the change in the depth with which students formulate their requests as an interaction develops may provide additional insight into their subsequent performance without AI.

Developing safe and useful general-purpose AI systems will require us to make progress on scalable oversight: the problem of supervising systems that potentially outperform us on most skills relevant to the task at hand. Empirical work on this problem is not straightforward, since we do not yet have systems that broadly exceed our abilities. This paper discusses one of the major ways we think about this problem, with a focus on ways it can be studied empirically. We first present an experimental design centered on tasks for which human specialists succeed but unaided humans and current general AI systems fail. We then present a proof-of-concept experiment meant to demonstrate a key feature of this experimental design and show its viability with two question-answering tasks: MMLU and time-limited QuALITY. On these tasks, we find that human participants who interact with an unreliable large-language-model dialog assistant through chat -- a trivial baseline strategy for scalable oversight -- substantially outperform both the model alone and their own unaided performance. These results are an encouraging sign that scalable oversight will be tractable to study with present models and bolster recent findings that large language models can productively assist humans with difficult tasks.

Concept Bottleneck Models (CBMs) are interpretable-by-design neural networks that detect human-understandable concepts from the input and use them to generate predictions. By allowing users to inspect the concepts underlying a prediction and explore how predictions change under alternative concept configurations, CBMs have emerged as one of the most prominent approaches to supporting human-AI collaboration. However, user studies investigating their actual effectiveness as decision-support systems remain limited. We present two large-scale user studies (N participants = 705, N observations = 6,959) evaluating how concept-based explanations and user interventions on the model's concepts affect the performance of the human-AI team in two distinct binary classification tasks. Our results show that CBMs, and particularly their interactive component, can improve human-AI team accuracy relative to both unaided human performance and performance with non-interpretable AI support. However, these benefits emerge only under certain conditions: classification tasks perceived as difficult, easily identifiable concepts, and active interaction with the model. We also discuss how inaccurate concept detection may undermine users' trust in the model. Overall, this work provides practical guidance for the deployment of CBMs as effective decision-support tools.

We introduce GDPval, a benchmark evaluating AI model capabilities on real-world economically valuable tasks. GDPval covers the majority of U.S. Bureau of Labor Statistics Work Activities for 44 occupations across the top 9 sectors contributing to U.S. GDP (Gross Domestic Product). Tasks are constructed from the representative work of industry professionals with an average of 14 years of experience. We find that frontier model performance on GDPval is improving roughly linearly over time, and that the current best frontier models are approaching industry experts in deliverable quality. We analyze the potential for frontier models, when paired with human oversight, to perform GDPval tasks cheaper and faster than unaided experts. We also demonstrate that increased reasoning effort, increased task context, and increased scaffolding improves model performance on GDPval. Finally, we open-source a gold subset of 220 tasks and provide a public automated grading service at evals.openai.com to facilitate future research in understanding real-world model capabilities.

Competitive music transcription models require large amounts of paired audio-score data, which is scarce due to collection costs, alignment difficulty, and copyright restrictions. Meanwhile, vast quantities of unpaired audio recordings and symbolic scores are freely available but have gone unused. We adopt a cycle-consistent translation framework in which a small amount of paired data acts as a minimal anchor, unlocking the full potential of the unpaired pool. We find that: unpaired data yields surprisingly large gains, especially under limited supervision; unpaired audio contributes more than unpaired scores; incorporating unlabeled audio from a new instrument during training improves transcription for that instrument without any paired supervision. Together, these results suggest that scaling unpaired data offers a practical path toward high-quality transcription for instruments where labeled data remains scarce.

Leaderboards rank long-horizon agents by their final outputs. Yet a higher score alone does not establish whether two systems are comparable or which stage accounts for the difference. Unequal evidence, inputs, or budgets can affect scores, and statistical corrections do not remove this mismatch. We introduce compositional controllability to address these questions. A comparison window covers one stage, several stages, or the whole agent. Our central result bounds the gap between observed and controlled score differences using only nuisance outside the window. This yields an admissibility test applied before scores are inspected. Inadmissible comparisons are refused. For admissible pairs, an ordering is certified only when the score gap exceeds the combined sampling and nuisance radii; otherwise, it remains undecided. These decisions give each system a rank interval. We introduce BioLitBench, a benchmark of 2,042 biomedical articles represented as structured claim graphs. Among seven published pipelines, a conventional statistical analysis declares a winner in 14 of 21 pairwise comparisons. Yet the top-ranked system alone received the target review's bibliography. To isolate pipeline performance, our test requires matched inputs and a fixed backbone model. It refuses 11 of the 21 comparisons, including every comparison involving the top-ranked system. Seven of the 14 conventional conclusions fall within these refused pairs. The same comparison windows support stage-level training. We train SCRIBE on Qwen3.8-27B using rewards measured at each stage's exit. Under matched evidence, SCRIBE achieves a certified rank interval of [1,2], with certified advantages over all evaluated published pipelines and the evaluated Claude and OpenAI agents. Under same pool, SCRIBE matches the strongest published retriever and is certified above three published pipelines.

Diffusion models have achieved remarkable success in text-to-image synthesis, largely attributed to the use of classifier-free guidance (CFG), which enables high-quality, condition-aligned image generation. CFG combines the conditional score (e.g., text-conditioned) with the unconditional score to control the output. However, the unconditional score is in charge of estimating the transition between manifolds of adjacent timesteps from x_t to x_{t-1}, which may inadvertently interfere with the trajectory toward the specific condition. In this work, we introduce a novel approach that leverages a geometric perspective on the unconditional score to enhance CFG performance when conditional scores are available. Specifically, we propose a method that filters the singular vectors of both conditional and unconditional scores using singular value decomposition. This filtering process aligns the unconditional score with the conditional score, thereby refining the sampling trajectory to stay closer to the manifold. Our approach improves image quality with negligible additional computation. We provide deeper insights into the score function behavior in diffusion models and present a practical technique for achieving more accurate and contextually coherent image synthesis.

When can an agent failure be caught? An audit is usually limited by the record rather than by the method. CatchBench therefore puts one auditor's question to three information states: the declared configuration before a run (PRE), a growing prefix of its trace (LIVE), and the finished trace (POST). Prior benchmarks fix one of these states or vary the telemetry; to our knowledge none scores all three under one task-method interface. Each state admits different questions, so seven task contracts carry their own labels and metrics rather than one leaderboard. Four are evidential; three are Gold-derived mechanism diagnostics. The release scores 72 entrants, from rule scanners and structural models to eleven LLM judges across nine model families (GPT, Claude, Gemini, Gemma, Llama, Qwen, DeepSeek, Mistral, Nova), over 1187 declared configurations and 1162 recorded runs. Most of the arena does not order: 56 of 138 registered contrasts separate, and the rest are published unresolved rather than ranked. The two sharpest results cut against our own data. One rule ignores every name and permission; it flags each capability declared after the first. On one of six configuration sources it reaches a perfect F1, so a score there measures how the corpus was built rather than how well a method reasons. Our admissibility bar then rejected one injected substrate and withheld evidential status from the other. A benchmark number is therefore not interpretable until the process behind its labels is published and tested for the shortcut it may leave. We report both, and regenerate every ordering from released predictions with no model call.

We introduce the first version of KWBench (Knowledge Work Bench), a benchmark for unprompted problem recognition in large language models: can an LLM identify a professional scenario before attempting to solve it. Existing frontier benchmarks have saturated, and most knowledge-work evaluations to date reduce to extraction or task completion against a specification. KWBench targets the step before that: recognizing the governing structure of the situation from raw inputs alone. The benchmark contains 223 tasks sourced from practitioners across acquisitions, contract negotiations, clinical pharmacy, organizational politics, fraud analysis, and incentive design. Each task encodes a formal game-theoretic pattern (principal-agent conflict, signaling, mechanism design failure, strategic omission, coalitional dynamics, strategic interdependence) and carries structured ground truth recording the expert reading of the situation and the anticipated failure modes. Models receive raw data and a task prompt with no indication of problem type. Scoring is a three-tier rubric gated by a mandatory conjunctive check. Mandatory criteria encode the predicted wrong paths. We evaluate 16 models. The best model passes on 27.9% of tasks. The top two models agree on only 31.7% of their passes. Among the top 8, 44 tasks are solved by exactly one model; routing across the top 8 covers 50.7% of the benchmark, nearly double the best single model. Conditional on passing, quality scores converge (approx 83% across models); unconditional scores do not. Same models articulate the relevant game-theoretic concept correctly when asked, then fail to apply it unprompted. We release KWBench to shift how frontier models are evaluated on knowledge work, scoring them on whether they recognize the right problem from the situation alone, not only on how well they execute once the problem has been framed for them.

Despite advances in Preference Alignment (PA) for Large Language Models (LLMs), mainstream methods like Reinforcement Learning with Human Feedback (RLHF) face notable challenges. These approaches require high-quality datasets of positive preference examples, which are costly to obtain and computationally intensive due to training instability, limiting their use in low-resource scenarios. LLM unlearning technique presents a promising alternative, by directly removing the influence of negative examples. However, current research has primarily focused on empirical validation, lacking systematic quantitative analysis. To bridge this gap, we propose a framework to explore the relationship between PA and LLM unlearning. Specifically, we introduce a bi-level optimization-based method to quantify the impact of unlearning specific negative examples on PA performance. Our analysis reveals that not all negative examples contribute equally to alignment improvement when unlearned, and the effect varies significantly across examples. Building on this insight, we pose a crucial question: how can we optimally select and weight negative examples for unlearning to maximize PA performance? To answer this, we propose a framework called Unlearning to Align (U2A), which leverages bi-level optimization to efficiently select and unlearn examples for optimal PA performance. We validate the proposed method through extensive experiments, with results confirming its effectiveness.

We study the problem of learning in zero-sum matrix games with repeated play and bandit feedback. Specifically, we focus on developing uncoupled algorithms that guarantee, without communication between players, the convergence of the last-iterate to a Nash equilibrium. Although the non-bandit case has been studied extensively, this setting has only been explored recently, with a bound of O(T^{-1/8}) on the exploitability gap. We show that, for uncoupled algorithms, guaranteeing convergence of the policy profiles to a Nash equilibrium is detrimental to the performance, with the best attainable rate being Ω(T^{-1/4}) in contrast to the usual Ω(T^{-1/2}) rate for convergence of the average iterates. We then propose two algorithms that achieve this optimal rate up to constant and logarithmic factors. The first algorithm leverages a straightforward trade-off between exploration and exploitation, while the second employs a regularization technique based on a two-step mirror descent approach.