Procedural mistake recognition usually gets tangled up in temporal segmentation and low-level vision features. This paper introduces PRISM, a language-grounded framework that maps video steps to text, optionally structures them with semantic roles, and decodes sequences with a lightweight hidden Markov model trained only on correct executions. By assuming benchmark-provided step boundaries, it isolates the core reasoning problem of detecting deviations in step content and order through a unified edit-based taxonomy.

The paper studies whether facts deleted from knowledge-graph embeddings stay deleted under ordinary continued training. It shows that standard approximate unlearners fail because entity embeddings are shared between forgotten and new facts, and proposes Kernel-Anticipated Unlearning (KAU), which uses the tangent kernel to predict and counteract this relearning inside the forgetting objective. This matters for the right to be forgotten in search, recommendation, and question-answering systems built on evolving knowledge graphs.

The paper tackles Zero-Shot Domain Generalization (ZSDG), where a model must recognize unseen classes in unseen target domains without access to any source-domain images during training. The proposed Domain Prompt Learning (DoP) built on CLIP introduces a two-stage pipeline: Text Domain Learning (TDL) synthesizes diverse style tokens via text-space orthogonality and consistency losses, and Visual Domain Learning (VDL) refines the best-matching token per test image using self-supervised entropy minimization at test time. This matters because it removes the costly and often infeasible requirement to collect and annotate source-domain data, making domain generalization practical for new or scarce classes.

Cinematic video generation demands both inferring plausible action and camera trajectories from sparse text or reference inputs, and independently controlling character, scene, action, and camera factors without unwanted side effects. The authors argue that simply feeding MLLM features into a Video DiT fails because the DiT can bypass those latents via native text-encoder and VAE-reference pathways. Factor-Aware Predictive Learning (FAPL) fixes this with a three-stage curriculum—latent-only grounding, factor-span replacement, and fixed-generator predictive learning—to force the DiT to rely on MLLM representations and, in turn, teach the MLLM to output generation-compatible latents.

Standard computer-model calibration focuses on parameter posteriors, not the predictive accuracy of the calibrated simulator. The authors propose a two-layer framework: a flow-matching model amortizes posterior inference, then conformal risk control tunes parameter sets so the simulator pushforward hits a target miscoverage on synthetic calibration data. This targets the quantity that actually matters downstream—predictive coverage—while exploiting amortization to avoid expensive per-observation MCMC. Experiments on synthetic and real epidemiological data show roughly 5× sharper bands than KOH–MCMC with near-nominal coverage.

Multilingual web-corpus pipelines such as FineWeb2 apply one recipe across thousands of languages, leaving individual languages’ failure modes unexamined. This paper proposes treating an LLM as a one-time debugger rather than a classifier: a budgeted, tool-using agent explores the Slovak slice of FineWeb2 through a read-only API and compiles each finding into a cheap executable test validated by a native speaker. The work matters because it replaces a linear, manual audit with a reusable test suite that scans 26.5M documents in under a second without a GPU.

World models are increasingly used to reduce environment interaction when post-training vision-language-action (VLA) policies, yet they are typically trained to minimize next-frame pixel error rather than to produce useful policy gradients. DRILL introduces a bilevel framework that trains the world model to maximize the policy's downstream return after an imagined GRPO update, instantiated as both a lightweight signed surrogate (DRILL-VWI) and a full meta-gradient method (DRILL-IMG). The work matters because it empirically decouples pixel fidelity from policy utility, showing that a world model with substantially worse pixel MSE can yield far better downstream policies.

Do people who speak different languages organize concepts in similar ways? The authors extract word embeddings from multilingual transformers (mBERT and XLM-R-large) to compare how 196 basic concepts relate to one another across 81 languages. Their core finding is that sociocultural meanings are more aligned across languages than nature and motion meanings, a domain-sensitive pattern that recurs across encoders and concept groupings. The work connects computational linguistics to cognitive science, though its inferences about human conceptual organization remain indirect because they rely entirely on neural network representations rather than behavioral or neural data from speakers.

Persistent web agents improve over repeated tasks only if they know user preferences that history alone cannot reveal. This paper formalizes Active Preference Acquisition (APA), where an agent decides which missing attributes to ask about by estimating their Prospective Personalization Value: the expected cumulative utility gain across future tasks minus asking and privacy costs. The core insight is that many preferences are worth acquiring only when reused across many later tasks, since no single booking or purchase may justify the cost of asking.

Monolingual language models predict native reading times well but miss how a reader's first language shapes second-language processing. This paper investigates whether bilingual GPT-2 models sequentially trained on an L1 and then English can capture human cross-linguistic influence during L2 reading. The authors evaluate 55 models---monolingual and bilingual with varying steps of L1 exposure---against eye-tracking data from 12 L1 backgrounds, showing that L1-matched bilingual surprisal improves prediction of late-stage reading measures and peaks for intermediate-proficiency learners. The work is significant because it provides a computational bridge between multilingual NLP and psycholinguistic theories of non-native sentence processing.

The paper tackles occupational activity recognition for construction workers using wearable IMUs. It shows that headline ~99% accuracies on the VTT-ConIoT benchmark collapse to ~77% mean under strict leave-one-subject-out evaluation, with a 27pp per-subject spread. The work systematically maps how sensor placement, window length, classifier choice, augmentation, and personalised calibration affect true subject-independent performance.

This survey tackles long-document visual question answering, where answers require locating and integrating evidence scattered across many pages and modalities including charts, tables, and fine-grained layout. Rather than grouping papers by model architecture, the authors decompose the task into four functional components: document parsing, evidence organization and retrieval, generator organization and learning, and evaluation. The core argument is that task bottlenecks—visual fidelity, evidence dispersion, and budget limits—should drive taxonomy design, not just technical novelty. This reframing matters because the field has grown explosively since 2024 and prior surveys have organized literature by approach rather than by the functional requirements that long documents impose.

Masked diffusion models freeze tokens once revealed, making decoding errors irreversible and bounding compute by sequence length. This paper reformulates MDM inference as a Markov chain over the clean sequence space via a remask-denoise kernel, derives a KL contraction bound that disentangles timestep schedule from model error, and proposes Random Remask-Denoise (R2D) inference and Trajectory Correction Training (TCT). The result is a reward-free, finetuning-free test-time scaling procedure that reports near-log-linear accuracy gains on hard Sudoku and 3-SAT and modest improvements on a coding benchmark.

AgentDreamer tackles text-controllable physics simulation by replacing ambiguous video diffusion priors with explicit 3D motion plans generated by an off-the-shelf multimodal agent. The agent reasons over scene geometry and text to produce dense parametric trajectories, which guide a differentiable MPM simulator toward physically feasible motion aligned with the instruction. A new metric, Motion-IG, isolates the information that simulated motion contributes to recovering masked text beyond the prompt alone. While the experimental scope is narrow, the framework offers a clean separation between semantic reasoning and physical grounding.

ReflectPO tackles autonomous error recovery in long-horizon LLM agents, where early execution mistakes compound across sequential tool-use interactions and existing RL methods optimize task outcomes without teaching the policy when and how to self-correct. The core idea is to generate recovery trajectories from rollback snapshots, guided by reflection hints derived from branch differences across rollout trees and repeated loops within trajectories, and then optimize the policy with tree-structured credit assignment that captures both local decision effects and recovery gains. This matters because preliminary evidence shows that errors are nearly ubiquitous in natural trajectories yet agents autonomously recover in fewer than 15% of cases, leaving substantial headroom for training-based correction.

Knowledge graph-to-text generation converts structured triples into fluent natural language, yet whether modern prompted LLMs can match task-specific fine-tuned models remains unclear. This paper benchmarks five recent model families—dense, mixture-of-experts, and encoder-decoder architectures—against a fine-tuned T5-large baseline on WebNLG and DART, using eight reference-based metrics together with a graph-grounded LLM judge validated by human annotation. The core finding is that prompted LLMs have nearly closed the gap on traditional metrics while exhibiting stronger faithfulness and fewer omissions on larger graphs, underscoring that surface similarity scores alone are no longer sufficient to assess generation quality.

Large language models can pick up hidden behavioral preferences—termed subliminal learning—when trained on data generated by other models. This paper investigates whether such transfer occurs in pretrained base models under hard-label and on-policy soft-label distillation, testing across different initializations, data orders, corpora, and architectures. By tracing how the abilities to transmit and receive a trait develop separately during pretraining, the authors challenge the notion that shared model origins are a prerequisite for unwanted preference transfer.

The paper attacks a structural restriction in the Bayesian Learning Rule (BLR): its reliance on KL geometry ties it to fixed-support exponential families, ruling out posteriors with adaptive tails or bounded support. The authors show that mixed-space Bregman discrepancies—specifically Fenchel-Young discrepancies—generalize the BLR while preserving its efficient additive canonical-coordinate updates. This matters because it yields a principled framework for scalable variational optimizers, such as $q$-VON, that use heavy-tailed or bounded-support $q$-Gaussian posteriors and can be implemented by modifying only a few constants in existing code.

Crisis warning generation is currently English-centric, hampering emergency response in linguistically diverse Europe. This paper introduces CrisiLangs, a multilingual dataset extending 1,000 expert-reviewed English event-message pairs into Italian, Bulgarian, German, and Greek via professional native speakers. The authors benchmark the resource by jointly fine-tuning Llama-3.1-8B-Instruct across all five languages, showing gains over zero-shot baselines but also revealing that generation quality and the utility of in-prompt guidelines vary substantially by language.

GM-GReFEL attacks two coupled problems in dynamic facial expression recognition (DFER): the quadratic cost of self-attention in long clips and noisy crowdsourced labels. It replaces a spatiotemporal Vision Transformer with a Modulated GroupMamba backbone that scans six channel groups—four spatially and two temporally—at linear cost $\mathcal{O}(N)$, re-mixing them via a Channel Affinity Modulation gate. A geometry-aware reliability module blends the classifier output with trainable facial-landmark anchors, weighted by per-sample normalized entropy, to fall back on anatomy when the prediction is uncertain. The authors report best-fold results within 0.3 points of the top method on DFEW and ahead on MAFW, while openly noting that five-fold means are more mixed.