Papers

Automatic academic paper-to-slide generation is inherently iterative, because creating an effective presentation requires repeated cycles of generation, critique, and revision. Recent multi-agent systems partially acknowledge this through internal critique-and-revise loops, while conversational approaches allow users to refine generated slide decks through dialog. However, these refinement processes either remain largely closed to the user or introduce feedback only after a complete deck has been produced, limiting the user's ability to participate in the iterative refinement of narrative flow, content allocation, and presentation emphasis. To address this gap, we introduce ConvDeck, a multi-agent pipeline for conversational paper-to-slide generation that distributes interaction across the pipeline through stage-specific loops, allowing users to iteratively refine both the presentation outline and the final slide deck at the stages where each kind of decision is made. These loops are driven by a refinement mechanism in which agents can think, speak, and act, enabling them to either directly apply edits or respond conversationally to clarify user feedback and discuss revision options. Our evaluation shows that stage-specific conversational feedback improves user-goal satisfaction while preserving narrative coherence, content quality, and visual presentation.

Weight-only post-training quantization (PTQ) can alleviate the computational burden of serving large language models (LLMs) at scale. However, existing PTQ methods often fail to generalize across models and suffer severe accuracy loss below 2 bits. Many leverage unstructured sparsity to mitigate this loss, but at the cost of regularity and GPU-friendly execution. We present QTEA, a sub-2-bit PTQ framework that quantizes weights into ternary values and uses salient weights as residual error compensators. To maintain hardware efficiency, residuals are assigned to selected columns with semi-structured $1:4$ sparsity within the salient columns. We further add column-wise rescale refinement to GPTQ-style column-by-column quantization, alternately updating per-column scales and ternary assignments to reduce reconstruction error. We also identify order-dependent error propagation in GPTQ and introduce error decay to attenuate late-stage error accumulation. On Qwen3-14B, QTEA compresses all weights to an effective 1.7 bits per weight while improving average accuracy over the strongest ternary PTQ baseline by 16.7%. It also achieves 1.40$\times$ and 2.61$\times$ lower perplexity on WikiText and C4 respectively. This trend holds on Llama3-8B, where QTEA obtains a 6.6% accuracy gain and 1.34$\times$ / 1.95$\times$ lower perplexity on the same datasets. Finally, we develop a lookup-table based kernel that achieves 7.2$\times$ faster per-token generation over an FP16 baseline. Code is available at https://github.com/Intelligent-Microsystems-Lab/QTEA.

Conversational artificial intelligence is increasingly embedded in everyday social environments, where it functions as both an informational tool and a source of interpersonal feedback. This perspective introduces contingency, i.e., the degree to which system responses vary with user behavior and its interpersonal consequences, as a central construct for evaluating AI systems. We argue that current alignment approaches, including reinforcement learning from human feedback, tend to prioritize user approval and conversational fluency over behaviorally informative feedback, leading to sycophantic patterns of noncontingent affirmation. Drawing on behavioral science and social learning theory, we propose that contingent feedback is a key mechanism through which individuals develop interpersonal skills. When AI systems provide feedback weakly coupled to social consequences, they may reduce opportunities for adaptive calibration in real-world interactions, particularly during adolescence, a critical period for social development. We outline a framework for contingent AI, including trajectory-based evaluation and models of social consequence prediction, and propose a research agenda spanning developmental psychology, human-AI interaction, and machine learning. More broadly, we argue that AI systems should be evaluated not only by user satisfaction, but by their impact on human social learning.

We seek to understand the effect of adding disruptive highly-capable new technologies to competitions by assessing the addition of Dynamite to Rock-Paper-Scissors. We find that providing a versatile Dynamite move to only one player provides limited value (win probability increases from 50% to 55.5%) and is played rarely. That value decreases further if the game is expanded beyond just the original three moves. We also observe several mechanisms by which prior moves can become strategically unplayable, or obsolete. We hope that this model illustrates some non-intuitive aspects of developing new versatile technologies. We also hope that it illustrates some pitfalls for developers and integrators to avoid in order to create value rather than merely capability.

Despite their growing use in video moderation, multimodal large language models (MLLMs) exhibit a compositional safety blind spot: videos composed of seemingly benign components can convey harmful meaning when interpreted as a whole. We refer to this phenomenon as Distributed Implicit Harm (DIH), where harm arises from relations among components distributed along a decomposition axis of the video, rather than from any single explicit cue. Among many possible axes, we study two representative cases: temporally distributed harm across visual segments (DIH-T) and cross-modal harm between audio and visual streams (DIH-M). Studying and mitigating DIH at scale requires data that is difficult to collect: such videos lack compositional harm annotations, evade retrieval based on local visual cues, keywords, or single-modality signals, and are consequently absent from existing safety datasets. To bridge this gap, we develop a multi-agent synthesis framework that composes individually benign components into harmful scenarios and generates diverse DIH videos with explicit reasoning annotations, yielding a dataset of over 9,000 videos spanning visual-only and audio-visual settings. Benchmarking over 30 MLLMs spanning frontier proprietary models and leading open-source systems reveals substantial and consistent deficits in detecting both DIH-T and DIH-M. Notably, this failure persists even among the strongest frontier models: they often correctly assess individual components in isolation but fail to recognize the harmful meaning that emerges from their composition. We further evaluate these models on a manually collected set of real-world DIH videos from social media and observe the same failure mode, highlighting DIH as a practical and underexplored challenge for video moderation.

Agent performance depends jointly on the model parameters and the executable harness code that manages context and control flow. Optimizing either component in isolation can leave the system bottlenecked by its frozen counterpart: weight updates can change which harness is effective, while harness updates can change which model capabilities are exposed. Existing joint-adaptation methods optimize weights and textual prompts but leave the broader harness fixed. We propose Weight-Harness Alternating LEarning (WHALE), a simple recipe that alternates two phases: updating the model under the current harness, then searching for a better harness under the updated model. We instantiate these two phases with online rejection-sampling fine-tuning and Meta-Harness, respectively. When to switch is a key design choice: to separate real improvements from noise without over-optimizing against a changing counterpart, WHALE uses either fixed phase durations or an adaptive patience rule over training signals. Using Qwen3.5-2B/4B agents across three domains (search question answering, mathematical reasoning, and chess puzzles), WHALE outperforms weight-only, harness-only, and Fast-Slow Training by 4.15-24.38 percentage points in best mean@8 accuracy. Either component can be the bottleneck: harness search matches peak weight-only accuracy with far fewer rollouts in SearchQA, but improves math accuracy only after a weight update. Small interleaved updates also outperform stagewise weight-then-harness optimization in accuracy and rollout cost. The code is available at https://github.com/krafton-ai/WHALE.

Document-to-slide generation is challenging because slides are dense editable artifacts that require both faithful content selection and precise spatial layout. Recent slide agents adopt iterative reflection, but typically follow a monolithic "one version, one feedback" loop: a slide or deck is rewritten, rendered afterward, and critiqued only at the turn boundary. This delayed feedback makes local failures such as overflow, overlap, clipping, and off-canvas placement difficult to attribute and repair. We propose ReDeck, a step-level render-grounded refinement framework that decomposes slide revision into atomic edit actions and returns renderer-derived observations after each step, turning refinement into "one edit, one observation." To balance local repair with global quality, ReDeck uses multi-granular feedback: step-level render feedback for spatial errors, a turn-level adaptive critic for semantic and design guidance, and a submission-level gate for hard layout validation. We further introduce DeckQuiz, a benchmark that decouples content fidelity, spatial correctness, and design quality. Across GPT-5.4, Claude-4.6, and Gemini-3.1, ReDeck consistently outperforms existing slide-generation agents, and ablations confirm that feedback timing and granularity are critical for reliable slide refinement.

We study federated online reinforcement learning with linear function approximation. While recent multi-agent reinforcement learning algorithms achieve strong regret guarantees, they typically require sharing raw trajectories. This reliance incurs a communication cost that scales linearly with the number of episodes and violates the privacy constraints of federated settings. To address these limitations, we propose Fed-LSVI, the first provably efficient federated algorithm for online reinforcement learning with linear function approximation in episodic Markov decision processes. By integrating a determinant-based event-triggered synchronization with a stepwise backward update mechanism, Fed-LSVI enables agents to collaboratively learn an optimal policy by exchanging only compressed sufficient statistics. We prove that Fed-LSVI achieves a regret bound of $\widetilde{\mathcal O}(\sqrt{Md^3H^4T})$, where $d$ is the feature dimension, $H$ is the horizon length, $M$ is the number of agents, and $T$ is the number of episodes per agent, matching the best-known regret for multi-agent online reinforcement learning with linear function approximation. Moreover, by following the stringent communication and privacy constraints of the federated setting, Fed-LSVI reduces the communication cost to only logarithmic dependence on $T$, representing a significant improvement over prior methods.

Public trust in Autonomous Vehicles (AVs) may depend not only on technical success but also on the fairness of their decision making. While a recent trend in AV research involves using general purpose "common sense" models to guide AV decision making, the degree to which these inherit human biases in driving is still understudied. Given that psychology studies have shown human driver biases exist, such as lower pedestrian-yielding rates to Black pedestrians in the US, we argue that analyses of model bias should also be part of AV evaluation. Concretely, in this paper we propose two new bias testing methodologies for Large Language Models (LLMs) and Visual-Language Models (VLMs), "All Else Being Equal" tests and "Self-Consistency" tests, and a new benchmark (FairYield) to assess bias in pedestrian-yielding decisions. Our findings show that the yielding decisions of both LLMs and VLMs can be influenced by models' predictions of pedestrian gender, ethnicity, religion, disability, age, skin tone, and socio-economic status. While the type and degree of bias is different from model to model, we highlight common patterns, and raise questions about the "common sense" model paradigm, particularly the need to either revise the paradigm or address issues of downstream bias.

Crisis helplines assess suicide risk through structured interviews, a process that is slow and dependent on operator training and workload. Natural language processing could support risk assessment and call prioritization, but almost no work addresses Arabic-language helpline calls or operates within the privacy constraints of real helpline data. We analysed de-identified transcripts from Lebanon's National Lifeline for Emotional Support and Suicide Prevention. Audio never left the helpline: calls were transcribed on site with a speech recognition model for Levantine Arabic, and an Arabic named-entity recognition model removed identifying information locally. Only the de-identified transcripts were shared with the research team. Operators recorded the five suicidal ideation items of the Columbia Suicide Severity Rating Scale, which we combined into two binary outcomes: at-risk and high-risk. We also machine-translated the transcripts into English, giving a paired Arabic/English comparison. On each corpus, we fine-tuned five instruction-tuned large language models alongside six transformer encoder baselines (four Arabic, two English) and evaluated all models on a held-out test set. We included 383 calls: 373 for the at-risk task (52.3% positive) and 297 for the high-risk task (30.0% positive). The best Arabic model reached a macro-F1 of 81.19 and a ROC-AUC of 90.61 on high-risk; the best English model reached 85.00 and 92.59, identifying 88.9% of high-risk calls. In both languages, high-risk calls separated more cleanly than at-risk calls, and translation to English did not reduce the best observed performance. Suicide risk can be classified from de-identified Arabic transcripts without sending audio outside the helpline. The high-risk results support further testing as an operator-facing tool; lower-severity ideation proved the harder case.

The number of edge devices in large-scale edge systems is rapidly increasing. Edge devices have limited processing power, memory, and network bandwidth, making resource utilization and data management during edge query processing challenging. Joins are among the costliest database operations in terms of time and resources. The State-of-the-Art edge query processing, Column Imprint-Hash Join CI-HJ, addresses this challenge using equi-height binning to accelerate hash joins. However, it lacks efficiency in real-time processing and scans unnecessary cachelines. This paper presents Workload Aware Column Imprint-Hash Join WACI-HJ, which uses a workload-aware approach to accelerate hash joins. Predicting the upcoming query workload in advance further improves its suitability for real-time edge query processing. WACI-HJ comprises two phases: WACI-HJ Generation Phase, including Pre-processing, Prediction, and Blocking and Hashing modules to compute bins based on the predicted workload before query arrival, and Query Processing and Resource Utilization, which handles query processing and CPU, RAM, and I/O utilization. Evaluations on a benchmark dataset and a real-world Smart Transportation dataset show a 54% reduction in cachelines read and 10% improved query execution time. The proposed technique is effective for both scaled and skewed data. Although PCR is an indirect measure of energy consumption, the work also directly measures energy consumption through energy-efficiency experiments. WACI-HJ shows 1%, 38%, and 49% gain in CPU, RAM, and I/O, respectively. Optimizing cache usage and query execution speeds up real-time traffic analysis, congestion management, and routing in Smart Transportation. Additionally, this technology can be applied to other domains to accelerate edge query processing.

Large language models sometimes deceive users without being instructed to. However, much of the study on deception in models involves instructed deception. We investigated the relationship between instructed and spontaneous (uninstructed) deception in Llama-3.1-70B-Instruct. We compared these two deception settings through direction geometry, cross-setting classifiers, and cross-setting steering. We found the two deception settings share a component of direction (cosine of approximately 0.5) and an asymmetry in the transfer between settings regarding detection and causation. Spontaneous trained classifiers performed better on instructed data than vice versa, and instructed derived directions performed better at steering spontaneous prompts than vice versa. Likewise the best token position to derive steering vectors from differed from the best token position to train and apply classifiers.

General-purpose NLP embedding models perform well on linguistic tasks, but their ability to capture symbolic ontological structure remains unclear. We introduce AVA, a systematic framework for evaluating whether embeddings distinguish logic-sensitive relational semantics in ontologies and knowledge graphs. AVA comprises 171,007 contrastive triplets derived from 163 heterogeneous ontologies using hierarchy inversion, relation substitution, and disjointness injection. Each triplet contains an ontology statement, a semantically equivalent paraphrase, and a logic-sensitive hard negative with contradictory relational meaning. We evaluate more than 25 state-of-the-art embedding models and find substantial limitations: the best model achieves only 0.739 triplet accuracy, while hard negative accuracy falls to 0.135. Fine-tuning improves discrimination by a large margin but transfers poorly to downstream Semantic Web tasks, including taxonomy discovery and ontology alignment. Further analysis suggests that improvements stem partly from perturbation-specific pattern recognition rather than robust ontological understanding. These findings reveal a persistent gap between linguistic representation learning and ontology-level discrimination, challenging the assumption that strong NLP benchmark performance translates to Semantic Web competence.

World models have made remarkable progress in action-conditioned future prediction for embodied agents, yet still struggle to model physically plausible interactions. Existing approaches address this limitation by constraining the generation process with external representations encoding motion, geometry, or semantics. Obtaining these spatiotemporally dense representations typically requires auxiliary estimators or manual annotations, limiting training scalability. We instead revisit the training objective and identify a supervision-allocation mismatch under the globally averaged mean squared error (MSE) denoising objective: prevalent static content dominates the optimization signal, leaving sparse dynamic-object regions critical to interaction generation disproportionately under-supervised. Motivated by this observation, we introduce IMPACT, a scalable Interaction-aware Model training framework with Prior-guided Attention Calibration and Targeting. IMPACT uses cross-attention associated with manipulated-object tokens as an internal spatiotemporal prior for action-conditioned changes. It samples candidate regions from this prior, calibrates them with detached local prediction errors to construct an interaction map, and uses the map to reweight denoising supervision, requiring neither external representations nor inference-time modifications. Extensive experiments on robot-arm and human-hand manipulation, spanning diverse control modalities and DiT backbones, show that IMPACT consistently outperforms the corresponding MSE-trained baselines, improving interaction fidelity, physical plausibility, and visual quality.

Latent language identification is often used to argue that multilingual language models route computation through language-specific states, such as English pivots. However, existing probes infer latent language from different signals, such as the geometry of hidden states or what can be decoded from intermediate representations. Since such claims shape conclusions about how models share and route information across languages, we ask whether these probes measure the same phenomenon or expose distinct aspects of multilingual computation. We study this question across model families, training regimes, domains, tasks, checkpoints, and up to 27 languages. We find that identification probes systematically disagree: the GMM-based representation probe, which draws evidence from hidden state geometry, shows earlier cross-lingual mixing, whereas decoding-based probes, which rely on output-space decodability, retain sharper language-specific and more English-biased signals. These differences track model multilinguality and training progression, but are comparatively stable across domains. Our results suggest a more cautious interpretation of latent language identification, where current probes expose different aspects of multilingual processing, rather than directly revealing a single internal lingua franca.

AI is increasingly used in the R&D process that produces future AI systems. We study the conditions under which this feedback becomes self-amplifying. Our model describes how the rate of AI capability growth depends on baseline research productivity, recursive feedback, and the increasing difficulty of research progress. We derive a recursive reproduction number, $\mathcal{R}_{\mathrm{AI}}$, that determines whether improvements are amplified or damped across development cycles. This quantity compares the strength of feedback with the rate at which further progress becomes more difficult. When $\mathcal{R}_{\mathrm{AI}}>1$, the effects of improvements compound across development cycles, placing the system in a self-amplifying regime. When $\mathcal{R}_{\mathrm{AI}}<1$, their effects weaken across cycles. The transition depends on the structure of the AI R&D feedback loop and need not occur at any particular level of model capability. A system can therefore enter a self-amplifying regime before acceleration becomes visible, while rapid progress can also occur without self-amplification. Higher baseline research productivity can accelerate progress without changing whether the system is self-amplifying, but the duration of the development cycle becomes a limiting timescale for amplification. Increasing research difficulty can end a period of self-amplification. Extending the model to multiple research actors shows that improvements shared across organizations can make the overall research ecosystem self-amplifying even when no individual actor is. The framework identifies measurable properties of AI R&D systems that can help distinguish recursive amplification from rapid progress driven by other sources, including the strength of recursive feedback, how effectively improvements propagate into successor systems, cycle duration, and the increasing difficulty of further progress.

The performance of artificial intelligence (AI) and machine learning (ML) models degrades when the problem they were trained on drifts. This is a near-universal feature of real-world problems, which often change unpredictably. Biological evolution has achieved intelligence by overcoming this obstacle through natural selection acting on heritable variation. AI/ML techniques have long incorporated forms of natural selection, but it has been challenging to maintain model diversity as optimization naturally drives convergence. Here we show that a swarm of AI/ML models subjected to deliberate mutations of their model coefficients away from optimality can reliably and sustainably improve performance in changing environments by acting as a statistical hedge against non-stationarity. We call this mechanism 'Flawed in Nature, Perfect through Evolution', reflecting that the collective performance gain goes at the expense of individual performance. We prove via four theorems that the resulting regret reduction is guaranteed under general conditions, establishing the Flawed-in-Nature mechanism as a generalizable design principle for AI/ML systems. We validate these results on synthetic linear regression tasks, demonstrating that the mutated swarm delivers the best model in $\sim80\%$ of environment changes and that inference synthesis successfully translates this individual advantage into a collective one. The mechanism proves to be most effective when the mutation drift rate matches the drift rate of the environment. We outline a simple, adaptive controller that enables practical applications by tuning the mutation drift rate to match the unknown drift rate of the environment. The close analogy of the Flawed-in-Nature mechanism to biological evolution suggests it may have been a critical missing ingredient for the organic discovery of AI forms that more closely mimic biological intelligence.

This paper presents FAIRY, a full-stack smart-agriculture agent system developed for and deployed to an operating soybean research farm at Harbin Institute of Technology's smart-agriculture site. We develop FAIRY to execute and evaluate agentic agronomic operations on full-season spatiotemporal workflows that span ridge preparation, planting, irrigation, fertilization, pest and disease treatment, harvest, grain handling, drying, and storage. FAIRY integrates APIs and infrastructure across production-grade machinery, fixed soil and canopy sensors, multispectral and thermal drones, satellite vegetation products, a weather station, calibrated crop-process models, agronomic records, and multi-season yield histories. The system is built around the novel "everything is an event" execution paradigm, which represents spatiotemporal world evolution, remote sensing and UAV observations, sensor readings, crop-growth transitions, machinery actions, and management interventions as state-changing events in a shared farm process engine. On top of this event-driven world model, FAIRY implements a complete agentic stack: a knowledge library of atomic agronomic skills; multi-agent controller and orchestration backends; frontier- and edge-model execution; full-path trace logging; and deployment profiling on local nodes. We use FAIRY to evaluate nine state-of-the-art agent controllers across one hundred full-season soybean scenarios that preserve the operational coupling between spatial observations in a 64-ridge field, temporal decision sequences, agronomic constraints, delayed effects, and final yield. We develop an evaluation suite that combines agentic success, full-path spatiotemporal correctness, token cost, and edge-device runtime.

Memory is widely viewed as an important unsolved problem for LLMs and VLMs, and current benchmarks typically evaluate it by testing accuracy over long text or video. However, accuracy alone misses properties that matter for real long-horizon tasks. We introduce ECCBench, a benchmark and evaluation protocol that measures memory beyond a system's capacity–its raw accuracy at a specific budget–via three axes we call ECC: efficiency–the computation, in FLOPs, needed to answer from memory; compression–whether compressible inputs are remembered more accurately or efficiently; and calibration–whether the system abstains in response to its own uncertainty and the cost of an error. We find that pretrained VLMs compress their memory over text but not video and are poorly calibrated on both. Among a broader set of memory backbones, several non-Transformer architectures achieve better compression-calibration tradeoffs than RoPE Transformers, suggesting they may be useful components for agents operating over long horizons.

Psychometric questionnaires contain rich item-level information, yet it remains unclear whether different representation learning objectives recover the same latent organization. We investigated this question using 757 matched teacher-child pairs from the baseline assessment of the Cyprus ProW preschool trial. Behavioral structure was characterized from child SDQ, ASBI, and CBRS item responses using principal component analysis and clustering, yielding four behavioral phenotypes. A contrastive objective substantially improved teacher-child retrieval relative to PCA-based representations, increasing Top-1 accuracy from 0.13% to 7.27% and Top-10 accuracy from 1.98% to 56.14%. However, contrastive representations preserved behavioral phenotype structure less effectively than PCA-based representations. A multi-task objective jointly optimizing alignment and behavioral prediction partially restored behavioral organization but reduced retrieval performance. These findings indicate that teacher-child correspondence and behavioral phenotypes represent distinct forms of latent organization and demonstrate that the latent structure recovered from linked psychometric data depends on the representation learning objective.

The development of long-context Large Language Models (LLMs) is constrained by the memory bandwidth bottleneck and quadratic complexity of the attention mechanism during decoding. To overcome the inherent trade-offs between the memory overhead of metadata-based metrics and the computational inefficiency of adaptive selection strategies, we present Faster Flash Decoding (FFD), a novel hardware-algorithm co-design framework designed to break the memory wall in long-context decoding. FFD integrates the selector and computer into a fully fused kernel, replacing external metadata indices with content-aware scanning via low-bit quantization. Furthermore, we introduce the top-delta strategy, which dynamically filters blocks to achieve distribution-adaptive sparsity without global synchronization. Offering a training-free and plug-and-play solution, FFD also enables the reuse of scanning results for computation, achieving up to 11.6x kernel-level speedup and scaling to 256K context length, with 2.37x end-to-end throughput improvement. Empirical validation on RULER and LongBench confirms that FFD maintains model accuracy while delivering high-ratio sparsity, with code available at https://github.com/qluoluo/faster-flash-decoding

Feature importance Methods (FIMs) are widely used in Explainable AI to interpret model predictions, yet attribution scores alone often provide limited insight into the underlying reasoning process. In this work, we introduce a novel perspective by embedding FIMs within a hypothesis-testing framework based on Weight of Evidence (WoE). We quantify how strongly the observed evidence supports any given hypothesis on feature importance. The reference hypothesis can stem from domain knowledge, ground truth, or be derived from the FIM itself. This formulation enables a principled evaluation of FIMs, capturing both their alignment with prior knowledge and their variability. We further provide theoretical results linking WoE to attribution variance. Empirical results shows the applicability and flexibility of our strategy analyzing LIME and SHAP explanations in settings with different reference hypotheses. Overall, our framework offers a complementary tool for assessing FIMs through a contrastive, evidence-based lens.

LLM judges, models that score another system's output, can be gamed by the systems they score. Recent work identifies one defence that works: the judge solves the task itself first and commits to that answer, then accepts a candidate only if the two match. We call this commit-first judging, and ask whether shipped software implements it, and what it costs. We audit the default judge configurations of eight widely used evaluation frameworks. Of the 24 configurations in scope, none implement it. Nine implement a variant the literature measures as ineffective, and share one ancestor prompt, traceable through a copied typographical error. In a controlled experiment, an ordinary best-of-N search with no access to correct answers optimises code against one of these configurations, used exactly as documented. On an interval merging task the judge accepted 90 of 96 candidates in one seed and 93 of 96 in the other; every accepted candidate passed every test the search could see and failed a held-out suite it could not. The judge identified the defective line and cited it as grounds for a perfect score. Commit-first judging removed the effect: 0 of 96 in both seeds. On a second task it made matters worse in both seeds: the judge's committed answer was wrong, and in one seed the population converged on it. This is our main finding. Commit-first judging does not remove the anchor that gets gamed, it moves it from the candidate to the judge's own answer, so evaluation is only as good as the judge is at the task. That precondition is cheap to measure in advance, and is task local rather than scale dependent: a smaller judge solved a task the frontier judge failed and resisted gaming where it did not. We also validate our own instruments: five of fifteen claims in our criteria were wrong against verbatim sources, and two held-out checks were unjustified by their specifications.

Large language models (LLMs) are increasingly used as rerankers in conversational recommender systems, yet measured gains depend strongly on the retrieval and inference protocol. On the ReDial conversational movie recommendation benchmark, we compare proprietary, open-weight, and fine-tuned LLM rerankers with collaborative-filtering and sequential baselines in a shared retrieve-then-rerank pipeline. We vary candidate-pool size, first-stage retriever, and decoding temperature. With a shared semantic top-250 candidate pool and strict candidate-aware scoring, the best proprietary reranker reaches NDCG@10 of 0.1497, compared with 0.0939 for the strongest non-LLM baseline. The same reranker reaches 0.2925 in zero-shot generation, showing that unconstrained scoring can yield a much larger apparent advantage than matched-pool evaluation. No evaluated open-weight LLM outperforms the tuned shallow autoencoder baseline under this protocol. For the strongest proprietary and open-weight rerankers, switching from semantic to collaborative-filtering candidates raises NDCG@10 by more than 50%, showing that measured reranker performance is highly sensitive to candidate generation. For the best proprietary reranker, raising temperature from 0 to 1.0 increases top-10 Jaccard distance from 0.0900 to 0.1240 while mean NDCG@10 changes negligibly, whereas weaker LLMs show larger degradation. These ReDial results support treating candidate generation, candidate-pool size, scoring policy, and decoding configuration as required reporting fields rather than implementation details.

LLMs acquire vast amounts of knowledge during pre-training, but often lack the specialized knowledge needed to answer questions from niche sources such as manuals or technical documents unseen during pre-training. Continued pre-training (CPT) is widely used to inject such knowledge into model parameters. However, niche documents seldom repeat facts, making it difficult for CPT to robustly acquire such knowledge. Recent works address this by generating multiple paraphrases of the new knowledge, but paraphrasing is computationally expensive and typically requires powerful LLMs. In this work, we introduce KItCAT: Knowledge Injection via Corrupted Auto-regressive Training, a lightweight training strategy that reduces the need for paraphrasing in decoder-only LLMs. KItCAT augments standard next-token prediction by stochastically corrupting the input sequence. During training, a random subset of input tokens is replaced with other vocabulary tokens while the original next-token labels are kept unchanged. This simple intervention generates diverse training inputs from each sample, enabling large-scale data augmentation at negligible cost. We show that KItCAT consistently improves over CPT across multiple datasets and model families. Code is available at https://github.com/meghanadhpulivarthi/KItCAT.

Parameter-efficient fine-tuning methods such as LoRA have become a standard approach for adapting large foundation models. Adopting fine-tuning to distributed settings faces several challenges. Most existing distributed LoRA methods rely on centralized aggregation, and gossip-based decentralized LoRA requires repeated synchronization among multiple model copies. Both methods incur significant communication overhead and introduce errors due to simultaneous aggregation of multiple model updates. In this paper, we take a different perspective and propose a random-walk-based LoRA fine-tuning scheme. Instead of maintaining multiple model replicas, a single model token traverses the network and is updated sequentially using local fine-tuning objectives. This design eliminates the need for global synchronization, substantially reduces communication and computation costs, and avoids aggregation errors. We provide rigorous convergence guarantees for non-convex objectives under standard assumptions. Through empirical results on multiple NLP tasks and graph topologies, we show that the proposed method achieves competitive task performance with substantially less communication and computation than gossip-based LoRA.

Clinical artificial intelligence is increasingly embedded in real-world care, yet existing safety mechanisms are poorly suited to reconstructing and learning from individual AI-related errors and near-misses. Aggregate model monitoring can identify performance changes, and traditional patient safety reporting can capture adverse events, but neither is designed to explain how risk emerges across the interaction among AI systems, clinicians, workflows, and institutional controls. We propose AI Morbidity and Mortality (AI M&M), a structured, blameless framework for case-based review of clinical AI failures. The framework combines standardized case intake, evidence preservation and investigator-level reconstruction, tool-in-loop attribution, and corrective-action tracking. Each event is classified across four linked dimensions: Trigger - Mechanism - Clinical Pathway - Corrective Action, separating the condition that exposed a vulnerability from the process that produced risk, its consequence for care, and the remediation assigned. We demonstrate the framework using five illustrative outpatient medication and clinical decision-support cases; two clinician reviewers independently applied all four classification axes and reached agreement across all 20 axis-level classifications. AI M&M is intended to complement, rather than replace, model monitoring, patient safety reporting, and regulatory oversight by converting individual AI-in-workflow failures into actionable institutional learning. Prospective evaluation across institutions, AI systems, and clinical settings is needed.

Malaria remains a significant global health burden, necessitating continuous research efforts to understand its complex molecular mechanisms, epidemiology, and potential therapeutic interventions. Extracting essential biomedical information from the vast and constantly growing malaria literature is a challenging task that demands innovative approaches. Recently, pre-trained language models have revolutionized natural language processing tasks, demonstrating remarkable capabilities in various domains. This paper proposes a fine-tuned pre-trained biomedical language model for biomedical information extraction from scientific literature on malaria disease. The proposed methodology selects and preprocesses a large corpus of scientific articles on malaria, and then annotates them with entities of clinical significance. It then leverages BioBERT, a state-of-the-art pre-trained language model, to encode the textual data into context-aware representations. We fine-tune the model using domain-specific annotations and supervised learning to enhance its ability to extract relevant biomedical named entities. Extensive experiments and comparisons with different encoding and machine learning algorithms show that the proposed approach significantly outperforms them in precision, recall, and accuracy. We also publish our human-labeled dataset for entity and relation extraction to enable other health informatics researchers to train advanced models for malaria information extraction.

Model-based control can directly execute specified objectives, while learning can amortize such behaviors into reactive policies, making their combination a natural solution to multi-stage manipulation. We introduce Semantically UNified (SUN) Programs, typed executables that compile grounded relations into aligned optimal control objectives, satisfaction predicates, and learning rewards. Our harness, Kuafu, equips a foundation model as a task-level agent to orchestrate scene preparation, verification, residual RL, and data production. The agent uses program feedback to repair candidate programs and training diagnostics to calibrate relative reward weights, retaining accepted task semantics across tool calls. Across nine multi-stage manipulation tasks, Kuafu achieves 82.03% average success rate, significantly outperforming all learned baselines. Its learned controllers generate demonstrations at 10.57x the human-teleoperation rate, yielding data that improve visualpolicy success by 23.6 percentage points over the strongest baseline. The policies transfer zero-shot to physical Franka and Kinova robots, demonstrating sim-to-real generalization.

The 2025–2026 AI market has seen a wave of stealth releases: frontier models launched anonymously on developer platforms under codenames. For their users, identity determines data-handling terms, supply-chain risk, and capability expectations. No validated methodology exists for black-box identity verification of anonymous models: practitioner checklists lack accuracy evidence, and self-identification is untrustworthy by design. We propose a four-stage forensic audit protocol for API-served models. Stage 0 reconstructs launch-time configuration from archived platform snapshots (Internet Archive), exposing preview–production drift. Stage 1 fingerprints configuration (context, output ceiling, reasoning, modality) against the platform catalog. Stage 2 tests tokenizer identity with a cross-length differential that rejects short-prompt collisions. Stage 3 corroborates with behavioral probes. We test declaration consistency on 10 known-identity releases (7 exact, 2 precision-differences, 1 partial, 0 counter-directional), not end-to-end identification under anonymity. Identification is validated prospectively on a flagship case whose 2026-08-23 analysis pointed to the GLM-5.3 version line and whose official reveal confirmed those family and version-line inferences (deployment variant was not pre-asserted; Flash was consistent post-reveal), and on three Stage-0-only cases where the protocol produced a graded hypothesis or declined rather than guessed. A standard-library-only implementation is provided as supplementary material.