Artificial Intelligence Papers

ALT World models let observers explore and interact with generated worlds. Recent work shows that when observers revisit a region, current models often fail to regenerate it consistently. We view this as a context retrieval problem. LLM agents manage context retrieval at two layers of their stack. At the model layer, the KV cache design determines how much context fits in memory. At the agent harness layer, harnesses such as Claude Code pick the context sent to the LLM. Prior work on world models has mostly focused only on the model layer. In this paper, we make three contributions. First, we build a benchmark to measure world model context retrieval. It generates new worlds for every episode, preventing a model from recalling rooms using training data. It also separates context retrieval from guesses based on nearby frames. Second, we build an agent harness that performs context compaction for world models, which to our knowledge no prior work has studied. Third, we use our benchmark to s

ALT Dungeons & Dragons (D&D) is a tabletop roleplaying game (TTRPG) in which a group of players, guided by a Game Master, roleplay a fantasy adventure with simulated combat and improvisational narrative. In D&D, an encounter is a single instance of simulated combat, where the players must use their characters to attempt to defeat a selection of enemy characters that the GM controls. D&D has multiple systems that approximate encounter difficulty by summing individual difficulty values that designers predefine for each enemy. However, since players and enemies can have a wide variety of capabilities, encounter difficulty depends on more than just the estimated difficulty of each monster. For example, these metrics do not account for positioning and dungeon layout, the synergies between monster abilities, and player party composition. We have created a tool, Mage-Hand, that can simulate D&D encounters semi-randomly using Foundry Virtual Tabletop, a virtual tabletop (VTT) system. In this paper

ALT Fine-grained audio-visual question answering (AVQA) requires models to resolve complex cross-modal correspondences and temporally grounding from sparsely distributed evidence. Omni-modal agents, which autonomously orchestrate perception tools to actively collect evidence, achieve strong performance, but it remains unclear whether active perception from both audio and visual modalities depends on broad specialized toolsets, especially audio or visual modality-specific tools. Such systems may also introduce long execution trajectories and substantial inference cost. We investigate whether comparable performance can be retained with a compact workflow exposing only essential perceptual capabilities. Starting from the specialized-tool OmniAgent, we conduct staged controlled experiments that prune the toolset and progressively consolidate modality-specific tools into omni-modal tools, motivating a two-tool agent comprising only Omni-modal captioning and Omni-modal QA. Across three widely us

ALT Organisations keep decades of financial records as scanned PDFs on shared drives, and AI agents can now reach those drives through Model Context Protocol (MCP) servers. Existing benchmarks do not mirror this production setting because they simplify the task. In some, the relevant documents are handed to the agent, so it never has to search. In others, the agent searches through documents with pre-extracted text, so it never has to read scanned images. We introduce TreasuryQA Bench, which requires both: 70 questions over 1,435 scanned U.S. Treasury publications, held in Veris's high-fidelity Dropbox twin. In each task the agent searches for the relevant file through Dropbox's MCP server, reads the pages it needs with Reducto's hosted OCR server or its own tools, and replies with one exact figure. A Python grader checks each answer against an accepted range around the correct value. In this paper we present results from ten agents. The best, Claude Opus 5.5 in Claude Code, solves 64.8% o

ALT Atmospheric chemical mechanisms are commonly distributed in the language of the Kinetic PreProcessor (KPP), which generates efficient Fortran or C code but does not fit into the differentiable programming frameworks in which machine-learning models are developed. We present KPPax, a kinetic preprocessor that generates differentiable, just-in-time compiled, and vectorisable chemistry code in JAX from mechanisms written in the KPP language. KPPax evaluates the rate law with sparse operations in log space, which keeps it differentiable, and integrates the stiff system with the implicit solvers and the forward- and reverse-mode adjoints of the diffrax library, so that exact sensitivities are obtained without hand-derived adjoint code. We demonstrate two applications with the POLLU benchmark mechanism. In the first, the Fisher information matrix computed from an automatically differentiated Jacobian predicts, before any fit, which of the 25 rate coefficients a flow-tube-like design can iden

ALT Language models pretrained on web corpora collected over multiple years learn both general and domain-specific knowledge. How these different kinds of knowledge are learned as pretraining progresses remains unclear. Recent point-in-time models trained on web snapshots from different years provide a way to probe how knowledge prevalence varies over time. Using these models, we propose a temporal variability score (TVS), defined as the coefficient of variation of an item's negative log-likelihood across models. A higher score suggests that the knowledge may be less consistently represented across years of web text. We evaluate six benchmarks covering diverse topics using six models from multiple open-source model families. We find that low- and high-TVS items are learned similarly early in training, but high-TVS items improve faster later on. At the end of the pretraining run, high-TVS items outperform low-TVS items in 34 of 36 model-benchmark settings, with an average advantage of 7.3 p

ALT When an LLM group's discussion is compressed into a short memory for a final decision maker, what should the memory keep? Two natural heuristics keep the rarest claims or the claims relevant to the decision criteria. On hidden-profile tasks both look attractive, because how widely a fact is held is confounded with whether it helps. We introduce CrossProfile, a diagnostic benchmark of 60 templates with automatically verified answer keys, in which the ownership of supporting, decoy, and irrelevant facts is crossed while their text is held fixed (480 tasks). Across Qwen2.5-7B, Llama-3.1-8B, and two DeepSeek models, both heuristics fail in predictable places. Selecting by relevance readmits decoys, because decoys are on topic: even a perfect relevance filter loses 23 to 31 points to rarest-first selection in the classic hidden-profile configuration. Rarest-first selection is no better than random selection on average and wins only where the rare facts happen to be the useful ones. In both

ALT We introduce Timestral, a family of four time series foundation models for probabilistic forecasting, with 5M, 28M, 123M, and 209M parameters. Timestral sets a new state of the art on GIFT-Eval, TIME, and fev-bench, where Timestral-209M and Timestral-123M take the top two places on every metric. We achieve this with three ingredients: Decoupled Horizon Forecasting (DHF), which separates prediction tokens from the causally encoded context, so that longer horizons are trained without masking context and each observed value is both a forecast target and context for later forecasts; a covariate framework that natively supports known-future covariates within a causal architecture; and a two-stage pipeline that pretrains exclusively on a diverse, large-scale synthetic corpus before post-training on a mix of real and synthetic data with extended context. This report describes the model design and its benchmark results. Weights for all four models are available on Hugging Face and inference co

ALT Many real-world decisions require teams to integrate information distributed across participants. Yet AI agents are still primarily trained and benchmarked in isolation. We introduce TeamGym, an environment for evaluating and training language model-based agents on team decision-making problems constructed from real-world legal, finance, medical, and software engineering corpora. TeamGym introduces a method to identify a minimal set of facts, each element of which is necessary and all of which are collectively sufficient for a correct final decision, which we call pivotal information. This enables TeamGym to both control the distribution of pivotal information and track how well agents identify and integrate pivotal information when making team decisions. Examining 8 language models on 12 source corpora, we find that current AI agents often fail to effectively share information: Distributing pivotal information across participants lowers success rate by 22.9 percentage points on averag

ALT GUI memory should preserve the results of past actions and learn reusable operations. This is difficult in long-horizon workflows, where agents must use information across applications and useful operations are interleaved with failed attempts. We introduce Task Advancement Memory (TAM), which defines task advancement as verified task progress, and supports ongoing task execution and learns reusable operations. During execution, Task Advancement State represents how verified task progress changes the remaining goals, allowing the agent to build on earlier achievements across applications. After each task, advancement supervision identifies and abstracts the interactions that produced a verified local result. The learned operations preserve the relation between required inputs, expected outcomes, and verification conditions. This retains locally successful operations from both successful and failed tasks. On MobileWorld GUI-only tasks, Advancement State alone raises success from 12.0% t

ALT Agents built on large language models are increasingly allowed to take consequential actions: cancelling bookings, issuing refunds, changing accounts. The rules that govern those actions, such as "obtain the customer's explicit confirmation before changing anything", are usually stated in the prompt and followed only as reliably as the model chooses to follow them. We present Concordat, a Python library that treats an agent as an object whose interface is a finite-state protocol, in the spirit of typestate and session types, and enforces that protocol on the model at run time. On every turn the model is offered only the tools that are legal in the current protocol state, calls outside that set are rejected before they have side effects, decisions reserved for humans can only be requested, and token and tool-call budgets are charged and split linearly across handoffs. Protocols are checked when they are defined, and a subclass of an agent is rejected if it could bypass its parent's prot

ALT Clinical research over electronic health records depends on derived variables whose definitions sit outside the schema, in item codes, time windows, missingness rules, and output-grain choices that change cohorts and downstream conclusions. We introduce M4Bench, 28 clinical derivation tasks across 15 task families on MIMIC-IV and eICU, to test whether AI coding agents can reproduce such derivations and how clinician-reviewed agent skills change their results. Its controls decompose the value of skills into specification completeness, delivery format, and schema familiarity. In a Docker-isolated Codex campaign (GPT 5.4 Mini and GPT 5.5), skills raise reward by 0.136 over the no-skill baseline, with 26 of 28 tasks improving. Specification completeness accounts for most of this gain: on an eight-task control subset, an operational spec alone raises reward by 0.162, and skills add only 0.024 on top, indistinguishable from zero. On a ten-task subset, decoy skills score like no skill and raw

ALT When a multimodal model computes 17+35 from text or 91-39 from a rendered image, it produces the same answer, 52. Does it use one internal concept of number for both, or separate machinery for each path? Similarity metrics such as RSA or CKA cannot settle this question. They give the same score to identical and to rotated directions, and two representations can look alike only because both prepare the same output digits. We therefore use causal interventions. For four conditions (text or image, addition or subtraction), we learn a small subspace of Gemma4-12B-it's residual stream in which swapping coordinates swaps the model's answer. The subspaces answer three questions. First, do the conditions write numbers along the same directions? They do not. Reusing a subspace from text in an image on the same operation recovers only 53.2% of its causal effect. A single rotation and rescaling maps one subspace onto the other and raises recovery to 91.4%, so the conditions share one geometry exp

ALT Agentic Entity Resolution (ER) systems iteratively plan actions, retrieve evidence, revise hypotheses, and propose graph mutations. An erroneous identity hypothesis may condition later retrieval and produce apparent confirmation. This paper proposes a Digital Identity Optimization (DIO) and Ontology of Digital Identity (ODI)-informed authorization boundary between evidence available for agent reasoning and evidence eligible to authorize persistent identity-changing mutations. A domain-governed constraint-binding function maps selected reconstruction-failure distinctions to explicit identity-preserving constraints. An abstract two-phase supervisor permits a commit only when claim-specific authorization-eligible evidence supports every required proposition, relevant conflicts are resolved, the operation remains within its authorized scope, and relevant state and policy versions remain valid. Admission as UNVERIFIED does not confer authorization eligibility. Separate claim-specific promot

ALT Large language models (LLMs) can answer many factual questions correctly, but a correct answer is not necessarily a supported one. For an LLM agent, a knowledge graph (KG) plays two roles: as a source of answers, it supplies facts that the model lacks; as a source of evidence, it supplies the supporting triples that make an answer traceable and auditable. However, existing KG agents treat the KG only as a source of answers: their training data carries answers but not evidence, and their training rewards check the answer but not the retrieved evidence, so agents often answer correctly without retrieving the supporting triples. We propose FoG, a framework that trains KG agents to retrieve the evidence behind their answers. FoG consists of evidence-grounded synthesis, a KG interaction agent, and an evidence-coverage process reward (ECPR): the synthesis recursively hides explicit entities behind verifiable sub-questions, so that every question carries its supporting triples by construction

ALT Generating several solutions can expose alternative methods, but additional trajectories often paraphrase one dominant approach. We investigate whether the conditional capacity of a frozen sparse mixture-of-experts (MoE) model can provide such diversity from one problem prompt, without user-assigned roles, prescribed methods, fine-tuning, or model copies. Agent-Conditioned Routing (ACR) assigns m parallel trajectories persistent, untrained shared/private expert-eligibility masks while preserving the learned, input-dependent token router. On a screened 126-question multi-mode mathematics and physics benchmark, with model judges reading the reported derivations in full, we count the number of problems yielding at least two distinct primary methods among soft-correct solutions in four trajectories. ACR reaches approximately twice as many such problems as the strongest temperature setting under GLM 5.3 (37.3 versus 18.7), and 1.67 times as many under GPT-5.6 Sol (45.00 versus 27.00), at ma

ALT Large language models (LLMs) are usually evaluated by final-output correctness or safety, yet an endpoint label does not reveal how a failure develops throughout the reasoning process. We therefore ask whether ordinary reasoning failures and adversarially induced failures can be characterized using the same internal coordinates. Towards this end, we in this paper study a white-box diagnostic framework CauTrace to disclose the root-cause of why LLM fails from three fundamental inspections: Functional States, Propagation of Representation, Temporal-State Coordination and Reasoning Trajectory. First, seven functional states are creatively defined to represent the support and constraint processing in LLMs. Second, the propagation of representation change measures how the changes among the seven states persist across subsequent Transformer blocks. Third, temporal-state coordination and trajectory topology characterize when the commitment and verification occur relative to the support, and h

ALT To be socially sustainable, Agentic AI systems require normative designs that align responsibilities with the ability to fulfil them. Identifying incongruences such as unsatisfiable expectations or privilege escalation is crucial in avoiding normative pitfalls. In this study, we discuss design dimensions suitable for describing multi-agent systems in light of this norm-ability alignment requirement. We propose a formal framework centred on process-oriented modelling and institutional power for the analysis of a given normatively relevant process, discuss strategies for repairing it, and elaborate on how the symmetrical relation between operational process and institutional state affects the lifecycle of normative positions.

ALT Statistical anomaly detection methods like Isolation Forest are efficient but suffer from a critical semantic gap. They identify numerical rarities but fail to understand the contextual meaning of data, often missing subtle, context-dependent anomalies. Conversely, Large Language Models possess strong semantic reasoning capabilities but are prohibitively expensive and slow for large-scale deployment. To bridge this gap, we propose ReForest, a novel framework that synergizes the statistical efficiency of Isolation Forest with the reasoning power of LLMs. ReForest transforms the latent decision paths of Isolation Forest into natural-language explanations, enabling an LLM to refine anomaly detection through an uncertainty-aware dynamic ensemble. This approach allows the model to adaptively weigh statistical and semantic signals based on uncertainty-driven instance-level confidence. Extensive experiments on four real-world datasets (Healthcare, Finance) demonstrate that ReForest significan

ALT Sequential recommendation predicts a user's next interaction from historical behavior. Recent LLM-based methods recast this task as generative recommendation and increasingly adopt reinforcement learning such as GRPO for post-training. However, these methods suffer from two limitations, namely credit-blind optimization, where a sequence-level reward is shared across all generated tokens, and collaborative-blind supervision, where each training signal is derived from a single user and ignores behavioral patterns shared across similar users. To address both issues, we propose Retrieval-Privileged on-policy Self-Distillation (RPSD). For credit-blind optimization, we introduce on-policy self-distillation (OPSD), in which a teacher with privileged knowledge re-evaluates the student's on-policy response, and the token-level confidence difference between the student and teacher reshapes the shared sequence-level advantage into token-specific credit. However, the privileged knowledge in vanill