The AI industry is constantly churning out news, like major acquisitions, indie developer successes, public outcry, and existentially dangerous contract negotiations.

Gavriel Cohen is living an open source developer's dream as his project has achieved acclaim and a partnership with Docker in a matter of weeks.

arXiv:2603.07311v1 Announce Type: new Abstract: This chapter explores the evolution of data-driven hint generation for intelligent tutoring systems (ITS). The Hint Factory and Interaction Networks have enabled the generation of next-step hints, waypoints, and strategic subgoals from historical student data. Data-driven techniques have also enabled systems to find the right time to provide hints. We explore further potential data-driven adaptations for problem solving based on behavioral problem solving data and the integration of Large Language Models (LLMs).

arXiv:2603.07316v1 Announce Type: new Abstract: While Large Language Models (LLMs) can accelerate text-heavy tasks in alternative investment due diligence, a gap remains in their ability to accurately extract and reason over structured tabular data from complex financial spreadsheets. Progress is held back by the lack of real industry fund portfolio datasets for benchmarking, as private equity data rooms are confidential. To address this, we introduce FinSheet-Bench, a benchmark of synthetic financial portfolio data modeled on real private equity fund structures, designed to evaluate LLM performance on text-serialized spreadsheet question answering and numeric reasoning tasks. Our evaluation of ten model configurations from OpenAI, Google, and Anthropic on financial spreadsheets, including complex layouts, fund dividers, and multi-line column names, reveals that no standalone model achieves error rates low enough for unsupervised use in professional finance applications. The best-performing model, Gemini 3.1 Pro, achieves 82.4% accuracy across twenty-four evaluation files of varying complexity and structural layout (approximately 1 error per 6 questions), followed by GPT-5.2 with reasoning at 80.4%, Claude Opus 4.6 with thinking at 80.2%, and Gemini 3 Pro at 80.2%. Performance degrades substantially on larger, more complex spreadsheets: the largest spreadsheet (152 companies, 8 funds) yields an average accuracy of just 48.6% across all models, compared to 86.2% on the easiest evaluation file. These difficulty patterns are consistent across all ten models, indicating that they reflect LLM limitations rather than idiosyncratic model weaknesses. Reliable financial spreadsheet extraction will likely require architectural approaches that separate document understanding from deterministic computation.

arXiv:2603.07335v1 Announce Type: new Abstract: High-performing vision language models still produce incorrect answers, yet their failure modes are often difficult to explain. To make model internals more accessible and enable systematic debugging, we introduce VisualScratchpad, an interactive interface for visual concept analysis during inference. We apply sparse autoencoders to the vision encoder and link the resulting visual concepts to text tokens via text-to-image attention, allowing us to examine which visual concepts are both captured by the vision encoder and utilized by the language model. VisualScratchpad also provides a token-latent heatmap view that suggests a sufficient set of latents for effective concept ablation in causal analysis. Through case studies, we reveal three underexplored failure modes: limited cross-modal alignment, misleading visual concepts, and unused hidden cues. Project page: https://hyesulim.github.io/visual_scratchpad_projectpage/

arXiv:2603.07369v1 Announce Type: new Abstract: How does the brain optimize sensory information for decision-making in new tasks? One hypothesis suggests learning reduces redundancy in neural representations to improve efficiency, while another, based on Bayesian inference, predicts learning increases redundancy by distributing information across neurons. We tested these hypotheses by tracking population responses in macaque cortical area V4 as monkeys learned visual discrimination tasks. We found strong support for the Bayesian predictions: task learning increased redundancy in neural responses over weeks of training and within single trials. This redundancy did not reduce information but instead increased the information carried by individual neurons. These insights suggest sensory processing in the brain reflects a generative rather than discriminative inference process.

arXiv:2603.07379v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) systems are increasingly evolving into agentic architectures where large language models autonomously coordinate multi-step reasoning, dynamic memory management, and iterative retrieval strategies. Despite rapid industrial adoption, current research lacks a systematic understanding of Agentic RAG as a sequential decision-making system, leading to highly fragmented architectures, inconsistent evaluation methodologies, and unresolved reliability risks. This Systematization of Knowledge (SoK) paper provides the first unified framework for understanding these autonomous systems. We formalize agentic retrieval-generation loops as finite-horizon partially observable Markov decision processes, explicitly modeling their control policies and state transitions. Building upon this formalization, we develop a comprehensive taxonomy and modular architectural decomposition that categorizes systems by their planning mechanisms, retrieval orchestration, memory paradigms, and tool-invocation behaviors. We further analyze the critical limitations of traditional static evaluation practices and identify severe systemic risks inherent to autonomous loops, including compounding hallucination propagation, memory poisoning, retrieval misalignment, and cascading tool-execution vulnerabilities. Finally, we outline key doctoral-scale research directions spanning stable adaptive retrieval, cost-aware orchestration, formal trajectory evaluation, and oversight mechanisms, providing a definitive roadmap for building reliable, controllable, and scalable agentic retrieval systems.

arXiv:2603.07427v1 Announce Type: new Abstract: As Large Language Models (LLMs) evolve into autonomous agents, existing safety evaluations face a fundamental trade-off: manual benchmarks are costly, while LLM-based simulators are scalable but suffer from logic hallucination. We present AutoControl Arena, an automated framework for frontier AI risk evaluation built on the principle of logic-narrative decoupling. By grounding deterministic state in executable code while delegating generative dynamics to LLMs, we mitigate hallucination while maintaining flexibility. This principle, instantiated through a three-agent framework, achieves over 98% end-to-end success and 60% human preference over existing simulators. To elicit latent risks, we vary environmental Stress and Temptation across X-Bench (70 scenarios, 7 risk categories). Evaluating 9 frontier models reveals: (1) Alignment Illusion: risk rates surge from 21.7% to 54.5% under pressure, with capable models showing disproportionately larger increases; (2) Scenario-Specific Safety Scaling: advanced reasoning improves robustness for direct harms but worsens it for gaming scenarios; and (3) Divergent Misalignment Patterns: weaker models cause non-malicious harm while stronger models develop strategic concealment.

arXiv:2603.07438v1 Announce Type: new Abstract: Regulatory stress testing requires projecting credit losses under hypothetical macroeconomic scenarios -- a fundamentally causal question typically treated as a prediction problem. We propose a framework for policy-path counterfactual inference in panels that transparently separates what can be learned from data from what requires assumptions about confounding. Our approach has four components: (i) observational identification of path-conditional means via iterated regression, enabling continuous macro-path contrasts without requiring a control group; (ii) causal set identification under bounded confounding, yielding sharp identified sets with interpretable breakdown values that communicate robustness in a single number; (iii) an oracle inequality showing that recursive rollout error is governed by a horizon-dependent amplification factor, providing a concrete answer to how far ahead one can reliably predict under stress; and (iv) importance-weighted conformal calibration bands with diagnostics that quantify extrapolation cost and trigger abstention when coverage guarantees degrade. The final output is a three-layer uncertainty decomposition that cleanly separates estimation uncertainty from confounding uncertainty. We validate all results through simulation and semi-synthetic experiments with real unemployment data, including a Covid retrospective demonstrating the framework's diagnostic value under extreme scenarios.

arXiv:2603.07462v1 Announce Type: new Abstract: Determining whether AI systems process information similarly to humans is central to cognitive science and trustworthy AI. While modern AI models match human accuracy on standard tasks, such parity does not guarantee that their underlying decision-making strategies are aligned with human information processing. Assessing performance using i) error alignment metrics to compare how humans and models fail, and ii) using distorted, or otherwise more challenging, stimuli, provides a viable pathway toward a finer characterization of model-human alignment. However, existing out-of-distribution (OOD) analyses for challenging stimuli are limited due to methodological choices: they define OOD shift relative to model training data or use arbitrary distortion-specific parameters with little correspondence to human perception, hindering principled comparisons. We propose a human-centred framework that redefines the degree of OOD as a spectrum of human perceptual difficulty. By quantifying how much a collection of stimuli deviates from an undistorted reference set based on human accuracy, we construct an OOD spectrum and identify four distinct regimes of perceptual challenge. This approach enables principled model-human comparisons at calibrated difficulty levels. We apply this framework to object recognition and reveal unique, regime-dependent model-human alignment rankings and profiles across deep learning architectures. Vision-language models are the most consistently human aligned across near- and far-OOD conditions, but CNNs are more aligned than ViTs for near-OOD and ViTs are more aligned than CNNs for far-OOD conditions. Our work demonstrates the critical importance of accounting for cross-condition differences such as perceptual difficulty for a principled assessment of model-human alignment.

arXiv:2603.07598v1 Announce Type: new Abstract: Chain-of-thought (CoT) improves reasoning reliability but increases token cost, motivating post-training compression of explicit reasoning traces. However, the shortest sufficient reasoning is not universal: it depends on difficulty, model capacity, and training state, making fixed length targets brittle. In practice, naive RL-based compression can also undesirably shorten the user-facing answer, because a single completion-level learning signal leaks across the think/answer boundary. We propose Difficulty-Scaled Segment-Wise GRPO (DSS-GRPO), which decomposes returns into think and answer components, computes group-relative advantages per segment, and routes them with hard token masks so compression updates act only on think while answer alignment acts only on answer. DSS-GRPO uses prompt-wise within-group shaping and difficulty-aware scaling to encourage concise reasoning without collapsing answer behavior.

arXiv:2603.07717v1 Announce Type: new Abstract: We test whether LLMs show robust decision biases. Treating models as participants in two-arm bandits, we ran 20000 trials per condition across four decoding configurations. Under symmetric rewards, models amplified positional order into stubborn one-arm policies. Under asymmetric rewards, they exploited rigidly yet underperformed an oracle and rarely re-checked. The observed patterns were consistent across manipulations of temperature and top-p, with top-k held at the provider default, indicating that the qualitative behaviours are robust to the two decoding knobs typically available to practitioners. Crucially, moving beyond descriptive metrics to computational modelling, a hierarchical Rescorla-Wagner-softmax fit revealed the underlying strategies: low learning rates and very high inverse temperatures, which together explain both noise-to-bias amplification and rigid exploitation. These results position minimal bandits as a tractable probe of LLM decision tendencies and motivate hypotheses about how such biases could shape human-AI interaction.

arXiv:2603.07728v1 Announce Type: new Abstract: Large language models (LLMs) such as GPT and Gemini have demonstrated remarkable capabilities in contextual understanding and reasoning. The strong performance of LLMs has sparked growing interest in leveraging them to automate tasks traditionally dependent on human expertise. Recently, LLMs have been integrated into intelligent agents capable of operating structural analysis software (e.g., OpenSees) to construct structural models and perform analyses. However, existing LLMs are limited in handling multi-step structural modeling due to frequent hallucinations and error accumulation during long-sequence operations. To this end, this study presents a novel multi-agent architecture to automate the structural modeling and analysis using OpenSeesPy. First, problem analysis and construction planning agents extract key parameters from user descriptions and formulate a stepwise modeling plan. Node and element agents then operate in parallel to assemble the frame geometry, followed by a load assignment agent. The resulting geometric and load information is translated into executable OpenSeesPy scripts by code translation agents. The proposed architecture is evaluated on a benchmark of 20 frame problems over ten repeated trials, achieving 100% accuracy in 18 cases and 90% in the remaining two. The architecture also significantly improves computational efficiency and demonstrates scalability to larger structural systems.

arXiv:2603.07733v1 Announce Type: new Abstract: This work investigated the capabilities of different models, including the Llama-3 series of models and CHATGPT, with different forms of expression in solving discrete optimization problems by testing natural language datasets. In contrast to formal datasets with a limited scope of parameters, our dataset included a variety of problem types in discrete optimization problems and featured a wide range of parameter magnitudes, including instances with large parameter sets, integrated with augmented data. It aimed to (1) provide an overview of LLMs' ability in large-scale problems, (2) offer suggestions to those who want to solve discrete optimization problems automatically, and (3) regard the performance as a benchmark for future research. These datasets included original, expanded and augmented datasets. Among these three datasets, the original and augmented ones aimed for evaluation while the expanded one may help finetune a new model. In the experiment, comparisons were made between strong and week models, CoT methods and No-CoT methods on various datasets. The result showed that stronger model performed better reasonably. Contrary to general agreement, it also showed that CoT technique was not always effective regarding the capability of models and disordered datasets improved performance of models on easy to-understand problems, even though they were sometimes with high variance, a manifestation of instability. Therefore, for those who seek to enhance the automatic resolution of discrete optimization problems, it is recommended to consult the results, including the line charts presented in the Appendix, as well as the conclusions drawn in this study for relevant suggestions.

arXiv:2603.07853v1 Announce Type: new Abstract: Research Agents enable models to gather information from the web using tools to answer user queries, requiring them to dynamically interleave internal reasoning with tool use. While such capabilities can in principle be learned via reinforcement learning with verifiable rewards (RLVR), we observe that agents often exhibit poor exploration behaviors, including premature termination and biased tool usage. As a result, RLVR alone yields limited improvements. We propose SynPlanResearch-R1, a framework that synthesizes tool-use trajectories that encourage deeper exploration to shape exploration during cold-start supervised fine-tuning, providing a strong initialization for subsequent RL. Across seven multi-hop and open-web benchmarks, \framework improves performance by up to 6.0% on Qwen3-8B and 5.8% on Qwen3-4B backbones respectively compared to SOTA baselines. Further analyses of tool-use patterns and training dynamics compared to baselines shed light on the factors underlying these gains. Our code is publicly available at https://github.com/HansiZeng/syn-plan-research.

arXiv:2603.07890v1 Announce Type: new Abstract: We propose image segmentation as a visual diagnostic testbed for coalition formation in hedonic games. Modeling pixels as agents on a graph, we study how a granularization parameter shapes equilibrium fragmentation and boundary structure. On the Weizmann single-object benchmark, we relate multi-coalition equilibria to binary protocols by measuring whether the converged coalitions overlap with a foreground ground-truth. We observe transitions from cohesive to fragmented yet recoverable equilibria, and finally to intrinsic failure under excessive fragmentation. Our core contribution links multi-agent systems with image segmentation by quantifying the impact of mechanism design parameters on equilibrium structures.

arXiv:2603.07972v1 Announce Type: new Abstract: While scaling individual Large Language Models (LLMs) has delivered remarkable progress, the next frontier lies in scaling collaboration through multi-agent systems (MAS). However, purely autonomous MAS remain ''closed-world'' systems, constrained by the static knowledge horizon of pre-trained models. This limitation makes them brittle on tasks requiring knowledge beyond training data, often leading to collective failure under novel challenges. To address this, we propose the Human-In-the-Loop Multi-Agent Collaboration (HILA) framework, a principled paradigm for human--agent collaboration. HILA trains agents to learn a metacognitive policy that governs when to solve problems autonomously and when to defer to a human expert. To operationalize this policy, we introduce Dual-Loop Policy Optimization, which disentangles immediate decision-making from long-term capability growth. The inner loop applies Group Relative Policy Optimization (GRPO) with a cost-aware reward to optimize deferral decisions, while the outer loop implements continual learning, transforming expert feedback into high-quality supervised signals that strengthen the agent's reasoning ability. Experiments on challenging mathematical and problem-solving benchmarks show that HILA, equipped with Dual-Loop Policy Optimization, consistently outperforms advanced MAS, establishing a principled foundation for collaborative and continually improving agentic systems.

arXiv:2603.07978v1 Announce Type: new Abstract: General-purpose computer-use agents have shown impressive performance across diverse digital environments. However, our new benchmark, OSExpert-Eval, indicates they remain far less helpful than human experts. Although inference-time scaling enables adaptation, these agents complete complex tasks inefficiently with degraded performance, transfer poorly to unseen UIs, and struggle with fine-grained action sequences. To solve the problem, we introduce a GUI-based depth-first search (GUI-DFS) exploration algorithm to comprehensively explore and verify an environment's unit functions. The agent then exploits compositionality between unit skills to self-construct a curriculum for composite tasks. To support fine-grained actions, we curate a database of action primitives for agents to discover during exploration; these are saved as a skill set once the exploration is complete. We use the learned skills to improve the agent's performance and efficiency by (1) enriching agents with ready-to-use procedural knowledge, allowing them to plan only once for long trajectories and generate accurate actions, and (2) enabling them to end inference-time scaling earlier by realizing their boundary of capabilities. Extensive experiments show that our environment-learned agent takes a meaningful step toward expert-level computer use, achieving a around 20 percent performance gain on OSExpert-Eval and closing the efficiency gap to humans by around 80 percent

arXiv:2603.07997v1 Announce Type: new Abstract: Although large language models (LLMs) are introduced into vision-and-language navigation (VLN) to improve instruction comprehension and generalization, existing LLM- based VLN lacks the ability to selectively recall and use relevant priori experiences to help navigation tasks, limiting their performance in long-horizon and unfamiliar scenarios. In this work, we propose CMMR-VLN (Continual Multimodal Memory Retrieval based VLN), a VLN framework that endows LLM agents with structured memory and reflection capabilities. Specifically, the CMMR-VLN constructs a multimodal experi- ence memory indexed by panoramic visual images and salient landmarks to retrieve relevant experiences during navigation, introduces a retrieved-augmented generation pipeline to mimick how experienced human navigators leverage priori knowledge, and incorporates a reflection-based memory update strategy that selectively stores complete successful paths and the key initial mistake in failure cases. Comprehensive tests illustrate average success rate improvements of 52.9%, 20.9% and 20.9%, and 200%, 50% and 50% over the NavGPT, the MapGPT, and the DiscussNav in simulation and real tests, respectively eluci- dating the great potential of the CMMR-VLN as a backbone VLN framework.

arXiv:2603.08013v1 Announce Type: new Abstract: Current Graphical User Interface (GUI) agents operate primarily under a reactive paradigm: a user must provide an explicit instruction for the agent to execute a task. However, an intelligent AI assistant should be proactive, which is capable of anticipating user intentions directly from continuous visual inputs, such as mobile or desktop screenshots, and offering timely recommendations without explicit user prompting. Transitioning to this proactive paradigm presents significant challenges. Real-world screen activity is rarely linear; it consists of long-horizon trajectories fraught with noisy browsing, meaningless actions, and multithreaded task-switching. To address this gap, we introduce PIRA-Bench (Proactive Intent Recommendation Agent Benchmark), a novel benchmark for evaluating multimodal large language models (MLLMs) on continuous, weakly-supervised visual inputs. Unlike reactive datasets, PIRA-Bench features complex trajectories with multiple interleaved intents and noisy segments with various user profile contexts, challenging agents to detect actionable events while fitting to user preferences. Furthermore, we propose the PIRF baseline, a memory-aware, state-tracking framework that empowers general MLLMs to manage multiple task threads and handle misleading visual inputs. PIRA-Bench serves as an initial step toward robust and proactive GUI-based personal assistants.