Computer Science & Mathematics

Detailed image captioning is usually framed as an amount problem: a model should mention more visual facts while avoiding hallucination. However, autoregressive VLMs generate captions as ordered sequences, and later generations may be produced under weaker or noisier visual grounding. We therefore argue that long-form captioning should be studied not only by what facts are mentioned, but also by where they appear. We introduce a position-aware rubric representation that decomposes dense human references into atomic visual claims, verifies generated captions against those claims, and induces a quartile-wise composition of supported, hallucinated, and generic content. This lens reveals two qualitatively different failures in long captions: models tend to produce unsupported specificity in the middle, while the final portion often escapes into vague, non-evidential language rather than simply accumulating more hallucination. A same-reference human control does not exhibit these patterns; additional human audits examine the visual validity of the induced labels. Motivated by this diagnosis, we propose PAGER (Position-Aware Grounded Evidence Reinforcement), a GRPO training objective with separate signals for supported fine-grained middle evidence, evidence-bearing tail continuation, claim coverage, and content-aware length control. The result is a closed loop: the same ordered claim representation identifies where long captions fail and supplies the position-specific reward needed to repair those failures. Experiments on detailed-caption benchmarks show that PAGER improves position-wise grounding and pairwise caption preference while preserving competitive global caption quality.

Recent advances in large language models and agentic workflows offer a new paradigm to early-stage structural design translating design intent into structural models. Although text-to-3D approaches have demonstrated that natural language can be used to produce visual, mesh-based geometric models, structural design requires domain-specific geometric arrangements and semantically rich data types. This paper introduces a proof-of-concept workflow for prompt-to-structure generation using an AI agent connected to a COMPAS-based structural modelling API through a Model Context Protocol (MCP) server. Through controlled modelling operations, the agent creates and modifies a graph-based structural data model composed of explicit building elements. The resulting artifacts can be inspected, modified, serialised, and further connected to downstream AEC workflows. The approach is demonstrated through selected prompt-to-structure examples and by progressive and editable design sequences that resemble typical iterative structural design workflows. The results are preliminary, but demonstrate the possibility of combining natural language, agentic workflows, and domain-specific structural data structures as a first step toward AI-assisted structural co-design.

In real-world applications, being able to explain why a prediction is made is often as important - if not more so - than achieving high accuracy. However, existing methods for sequential event prediction typically prioritize accuracy at the expense of interpretability. Neuro-symbolic approaches offer a promising direction to improve interpretability by learning symbolic rules, but extending them to next-event prediction is challenging due to the presence of temporal dependencies. We introduce RULES, a neuro-symbolic model that learns sequential rules directly from data, and combine it with a residual network to form RUNES (Rule Network for Event Sequences). RULES accurately recovers ground truth rules from synthetic data and discovers compact, interpretable rule sets on real data. RUNES combines these rules additively with a residual network, achieving competitive accuracy while providing transparent access to active rules and how residuals adjust predictions.

This paper presents a dataset of 121,536 ground-reasons pairs across 86,464 rulings of the five civil chambers of the French Court of Cassation (2016-2025). An appeal on points of law cannot itself be appealed and often decides a case for good. Understanding how the Court rules is thus crucial and can help lawyers who draft those appeals. Inside a ruling, every ground of appeal (a distinct legal complaint against the appealed decision) has its own answer from the Court. This answer is called the reasons and it belongs to one of a few doctrinal families. Using a deterministic pipeline, we pair each ground with its reasons and attach the pair to its family, according to the established French taxonomy. We obtain a descriptive grammar of civil cassation that shows which grounds succeed, which legal texts they refer to, and how both evolve over the decade. We release the dataset, the code and an interactive consultation tool.

Procedural mistake recognition requires identifying deviations from valid multi-step activities. We study post-completion recognition from videos with benchmark-provided step boundaries. We introduce PRISM (Procedural mistake Recognition via Interpretable Semantic Modeling), which grounds video segments in language and maps their descriptions to procedural states using a lightweight hidden Markov model trained on correct executions. Reference alignment and semantic checks identify errors in step content and order. To support evaluation across heterogeneous benchmarks, we map dataset-specific annotations to four edit-based error types: insertion, deletion, substitution, and transposition. Across three benchmarks, we compare gold and VLM-generated descriptions in raw, semantic-role, and action-object forms. Decoder controls show consistent gains from the trained local decoder over nearest-candidate matching, with additional, dataset-dependent benefits from sequence decoding. Semantic role labeling improves Any-F1 in all six matched grounded comparisons, although the gains vary across datasets and do not translate into uniform improvements under separate configuration selection. These results establish language grounding as an effective interface for explicit procedural mistake reasoning and characterize the contributions of representation and decoding.

Frontend coding agents are usually graded on one question: did the requested target get built. Real engineering changes carry a second obligation, because the rest of the site has to keep working. We call that second obligation website continuity, and introduce ProductWebBench, a benchmark that scores both on living frontend repositories: 320 Change tasks over existing code and 80 long-horizon Build tasks. Every task freezes a repository snapshot, staged requirements, browser evidence and deterministic checks before evaluation, and enters only once a reference passes every check and plausible-wrong ones fail. A simulated user reveals the requirements one stage at a time, and every check cleared earlier stays armed for the rest of the run. Across 8,737 runs from 13 models the two come apart: of the Change runs passing every requested-work check, 28.2% still fail a continuity check -- a blank route, horizontal overflow, console errors -- and 10.8% push a browser state they had already passed back to failing. Because the same checks also run on the untouched repository, each regression carries exact provenance, not a post-hoc label: on the 586 runs where both readings exist, 18.3% are caught breaking a verified state. Interaction is a further blind spot, failing in 29.5% of runs whose static DOM already passes. ProductWebBench therefore scores a patch against the living system it lands in, not the target alone.

We focus on optimisation-based Decision Support Systems in real-world applications. In this context, we aim to provide contrastive explanations that answer users' questions about recommendations based on the solution of an optimisation problem, specifically a Mixed-Integer Program (MIP). Addressing this endeavour is crucial to ensure human oversight of the decisions implemented, as required by international guidelines on trustworthy and responsible automated decision-making systems. We propose a novel methodology that combines constraint reasoning techniques and Large Language Models (LLMs) to generate contrastive explanations for optimisation problems encoded as MIPs. Although LLMs have shown impressive capabilities in Natural Language processing, they can produce incorrect answers due to reasoning limitations. To overcome these limitations, we leverage constraint reasoning techniques to distil the information necessary to produce an explanation, which is then passed as input to an LLM as the so-called graph of reasons. Such a graph captures the semantics and the relationships among constraints of the optimisation problem. Results on real-world scheduling domains show that incorporating constraint reasoning improves explanations generated by the LLM. In particular, using the graph of reasons enables the LLM to produce natural-language explanations that include more relevant reasons while reducing errors compared to explanations generated without this structure.

Diffusion language models generate multiple token positions in parallel, but few-step decoding makes parallel commitment difficult: tokens that appear reliable individually may be jointly inconsistent, while premature commitments can prevent later context-dependent refinement. In this paper, we introduce SetCommit, a training-free method for coordinated parallel token commitment. Instead of selecting tokens independently by confidence, SetCommit selects a commitment set that balances token reliability against local joint commitment risk, preserving conditioning opportunities for subsequent denoising steps. It further performs budget-neutral local exchanges, replacing vulnerable earlier commitments when newly revealed context supports more reliable alternatives. SetCommit requires no retraining or architectural modification and is compatible with existing diffusion decoding strategies. Experiments with LLaDA-8B and Dream-7B on GSM8K, MATH-500, MBPP, and HumanEval show accuracy/pass@1 gains of up to 7.4 percentage points over native decoding at 32 model evaluations. At common quality targets, geometric-mean incremental speedups range from $1.44$ to $1.74\times$. Code is available at https://github.com/Kissmetothemoon/SetCommit.

Generated robot videos can look convincing while depicting the wrong interaction: a different object is moved, a grasp disappears, or a placement ends outside its destination. We introduce RGraph-Eval, a reference-based framework that makes these discrepancies explicit through task-grounded temporal scene graphs. Unlike approaches that construct separate evidence representations for different evaluation dimensions, we organize task entities, interactions, and motion within a unified, interpretable scene graph representation. A foundation-model pipeline identifies manipulation roles, grounds persistent object masks, infers and verifies state transitions and action intervals, and reconstructs manipulated-object motion in three dimensions. Applying this representation to a generated video and a reference demonstration supports an interpretable RGraph-Eval score with visible-object, relation-change, and 3D path-shape components. We also introduce ShareRobot-SG, a collection of 1,000 manipulation scenes from ten ShareRobot sources annotated with the multimodal representation. Our evaluation separates graph-generation quality from agreement with human video judgments and includes comparisons with adapted SVG2 and SG-Ego, as well as a benchmark of three video generators. The framework connects robot-video assessment to the observable entities and interactions underlying a task while making the graph annotations reusable beyond evaluation. Code will be publicly available at https://github.com/SashaDance/rgraph-eval.

Measuring animal distance and motion in metric units underpins abundance estimation, morphometrics, and movement ecology, yet monocular video lacks metric scale and wildlife datasets lack metric ground truth. We present WildDepth 2, extending the synchronized RGB-LiDAR WildDepth dataset in three directions. (i) Imaging radar: 37 RGB-LiDAR-4D-radar sequences (19,716 frames) with per-point Doppler velocity and a measured error budget, to our knowledge the first imaging-radar wildlife recordings. (ii) Benchmarking: a frozen, checksummed hidden test set for RGB-only distance estimation and RGB+sparse-LiDAR depth completion, evaluated separately on animal and background pixels. (iii) Motion and tracking: metric 3D trajectories obtained by lifting 2D tracks using LiDAR, radar, or monocular depth. Across 48,723 animal frames, object-level radar calibration improves monocular depth over scene-level LiDAR calibration without requiring LiDAR at inference (AbsRel 0.120 vs. 0.198), although animal-level LiDAR remains more accurate (0.052). Across 431 tracks, once-per-track radar scaling reduces 3D position error four-fold (2.20 to 0.58 m) and distance-sampling detection-function bias from -13% to -2%, matching once-per-track LiDAR calibration. Transferring radar Doppler onto LiDAR produces motion masks 3.7 times denser than radar alone at 72% precision, reduces moving-animal LiDAR accumulation error by 37%, and estimates radial speed (r = 0.60), albeit with approximately 50% underestimation.

Our previous analysis found that a model's native hashed N-gram table is where continual updates disturb the base least, but its writes needed backpropagation and changed the pretrained table. We present Delta N-gram, a forward-only write path into that table. The pretrained table T^0 is never modified. Evidence is written into a sparse patch Δon the rows it addresses: a 4.1M-parameter writer emits a target and a gate per (position, hash head) event, and a block-averaged delta rule keeps each row a convex combination of its old value and the gated targets. The writer is trained offline on raw text with the pretraining objective; at use time writing takes forward passes only, and clearing Δrestores the original model. A 0.6B prototype trained to answer questions keeps 16/16 facts after a second write; online gradient writes keep 2--4/16. We study two writers: a plain writer that encodes evidence with the T^0-only model, and a delta writer that reads its current memory while encoding. On Qwen3.8-Flash-Next, the plain writer's memory is document-specific: on 64 held-out news events it lowers continuation NLL by 0.068 nats/token against empty memory and by 0.024 against another document's memory, and 0.063 and 0.013 after 16 intervening articles. The delta writer's memory is more document-specific and keeps more of it under distraction (70% vs 54%). Over 35,000 writes on our previous seven-task chain both writers stay stable and the delta writer drifts less (layer-2 math drift 176% vs 188%); later accuracy changes by -0.26/-0.52pp (-7.29pp for gradient updates) and small-k_i drift is 17.9/17.5% (39.2%). Acquisition is negative, and general-input drift reaches 32.8% (20.75%): format rows read by every example grow to 4--12 times their pretrained norm. Zeroing eight such rows removes over 98% of layer-2 drift on math inputs, together with their NLL gain.

Formulated a hundred years ago, the Bridgman formula for thermal conductivity of liquids postulates a proportionality between sound velocity and thermal conductivity. This designates phonons as heat carriers in a liquid. Its empirical success has been under-appreciated and its conceptual implications overlooked. It is not mentioned in well-known condensed-matter textbook. Here, we review its origin, its links to hydrodynamic Euler equations and to the fundamental paradox surrounding the parabolic heat equation and its fate across the quantum-classical crossover.

Offline tests can fail to reveal whether an agent update will remain safe after deployment. We operationalize this post-audit mirage using matched safe and harmful lifecycles that expose identical verifier-visible evidence while an isolated scorer records opposite deployment truth. Our theorem applies standard observational-equivalence and indistinguishability ideas to this executable audit boundary and states the resulting error-abstention tradeoff. The Post-Audit Mirage benchmark makes that boundary executable across authorization, scheduling, and shared-capacity environments. IdentifiedRangeMonitor uses fresh group-labelled harm outcomes to return deploy, hold, cannot determine, or unsupported. Experiments show that monitoring can recover the missing distinction in favorable settings but may remain undecided for rare groups, small effects, or missing observations. Verifier-agent evaluation must therefore test both calculation validity and whether the available evidence supports the deployment decision.

Real-world deployment of personalized federated learning (PFL) requires adapting to heterogeneous clients while preserving recognition performance beyond local distributions. However, high local accuracy can conceal failures on locally rare or absent classes. Effective personalization should therefore strengthen global recognition while raising the personalization floor. We propose an evaluation protocol that complements local accuracy with global, rare-class, and unseen-class evaluations to reveal generalization obscured by local scores, and assesses personalization floors among the worst-performing clients to expose weaknesses hidden by aggregate performance. To meet these deployment challenges, we propose FedGRC, a unified PFL framework integrating generalization and personalization through (i) Consensus-Guided Representation Learning (CGRL), which learns discriminative representations through shared directional and magnitude guidance; (ii) Multilayer Metric Synthesis (MMS), which constructs a shared covariance metric from these multilayer representations for statistical classification; and (iii) Support-Adaptive Shrinkage Personalization (SASP), which adapts class means within this metric by shrinking local estimates toward shared means according to local sample support, enabling adaptation while retaining shared knowledge. Extensive experiments, including benchmarking against 21 methods across 5 heterogeneity settings on 3 datasets and comprehensive ablation studies, validate the effectiveness of FedGRC. On CIFAR-10 and CIFAR-100, FedGRC surpasses state-of-the-art (SOTA) baselines in global recognition and personalization floors by at least 2.17% and 2.13%, respectively. On the real-world heterogeneous FedRS dataset, it outperforms the respective best baselines by 2.15%, 3.30%, and 4.61% in local adaptation, global recognition, and the personalization floor.

Long-context models often use answer-bearing evidence less effectively when it appears away from a prompt boundary. We compare this position bias across Transformer, state-space, and hybrid Mamba architectures by moving one gold document through a fixed set of nine distractors. The evaluated Transformers show a beginning advantage that the evaluated state-space models lack, while both families favor evidence near the end. A matched pure-versus-hybrid Mamba comparison points in the same direction but remains statistically uncertain. The family difference appears among models that can answer the task. Transformer primacy reappears on synthetic retrieval, where state-space ceiling prevents a family comparison. A corpus swap within one state-space architecture changes overall accuracy without a detectable change in curve shape. Prompt changes can create or nearly remove a beginning advantage. Position remains decodable across architectures even when response curves differ, and attention concentration near initial tokens covaries with beginning sensitivity in the tested Transformer family. These exploratory results show that effective context use depends on the evaluated architecture family, capability, task, and prompt, without isolating the sequence mixer as the sole cause.