Translate
Live intelligence

Latest AI updates

1,700 updates
AllModelsProductsIndustryResearchTutorialOpinion
Research Editor selected

A.S.E 2.0: How Repository Context Impacts the Security of LLM-Generated Code

This paper introduces CASE-Bench, a benchmark of 145 repository-level tasks derived from real CVEs across 51 projects, to study how repository context affects the security of LLM-generated code. Experiments with 13 context construction methodshol and six LLMs show that functionality-oriented context improves functional pass rates more than secure pass rates, and that withholding safeguards or inserting insecure examples significantly increases exploitable code. Coding agents observe relevant safeguards in only 27-36% of trajectories, highlighting the need for security-aware context construction.

Why it matters: Provides concrete, recent evidence and a new benchmark on a critical but understudied aspect of LLM code security, offering actionable insights for developers and researchers.

arxiv1d agoScore 85
LLM code securityrepository contextsecure code generation
Research

Multi-Resident Data Association in Smart Homes via MDP Modeling and Behavior Cloning

This paper introduces a novel framework for multi-resident data association in smart homes, reformulating it as a Markov Decision Process solved via behavior cloning. The approach uses a lightweight Transformer and a fine-tuned LLM, outperforming state-of-the-art baselines on three datasets, especially for concurrent events. The LLM shows preliminary cross-environment transfer potential, though real-world robustness needs further validation.

arxiv1d agoScore 78
multi-resident activity recognitionambient sensorssequential decision making
Research

STEK: Structural-Terrain-Enhanced Koopman-Inspired Cross-Channel Regularization for Retrospective Battery Multi-Target Estimation

The paper introduces STEK, a framework for retrospective battery multi-target estimation that combines cell-level lifecycle dQ/dV terrain with endpoint-derived structural proxies via quality-proxy-conditioned gating. It uses bounded FiLM and a validity-gated residual head to refine endpoint-curve representations, plus a Koopman-inspired low-rank linear map for cross-channel consistency. Experiments under the MIX-battery protocol show improvements over endpoint-only baselines and direct concatenation, with graceful fallback when descriptors are removed.

arxiv1d agoScore 72
battery health estimationmachine learningKoopman theory
Research

Engineering Bounded Halting and Observable Universality Tests

This paper reframes the halting problem as a bounded-halting decision with a step budget Taine, and shows that no deterministic checker can decide all instances within a fixed fraction of T below one. It also proposes adversarial tests for finitary universality that check resource growth and absence of fixed deliberation bounds, with implications for machine-learning systems.

arxiv1d agoScore 72
halting problemcomputational complexityuniversal machines
Research

Reproduction Study Finds Risk-Controlled Throughput Forecasting Fails Under Temporal Splits and Low Budgets

A reproduction of the BG-CFQS method for risk-budgeted LEO satellite throughput forecasting shows that its risk guarantee holds only under random calibration splits, not the contiguous temporal splits seen in deployment. The method cannot serve risk budgets below 0.15, overshooting by up to 3.7x at the tightest budget. Online calibration tracks budgets well, but group-conditional recalibration often underperforms simple conditioning.

arxiv1d agoScore 78
risk-controlled forecastingLEO satellite networksconformal prediction
Research

Conformal Prediction Framework for Distribution-to-Distribution Regression

Researchers propose a new conformal prediction framework for distribution-to-distribution regression, a setting where both inputs and outputs are probability distributions. They show that naive Split Conformal Prediction produces prediction sets that are marginally valid but lack interpretable volume or tractable sampling. Their approach operates in an orthogonal basis space to yield interpretable prediction sets while preserving coverage guarantees, with an adaptive variant giving asymptotic conditional coverage. Empirical results on synthetic and real-world datasets validate the method, and code is publicly released.

arxiv1d agoScore 72
conformal predictiondistribution-to-distribution regressionuncertainty quantification
Research Editor selected

EnzymeRSI: A dual-loop recursive self-improvement system for enzyme design

EnzymeRSI is a dual-loop recursive self-improvement system that evolves harness assets and algorithms without changing model weights, closing the loop with a cloud laboratory. It improved ProteinGym Spearman correlations by 0.044 on average, reached a PCC of 0.849 on ThermoMPNN, and demonstrated wet-lab success including a 31.1-fold catalytic efficiency increase in KOD-E polymerase. The system offers a new paradigm for autonomous enzyme engineering by treating workflow and algorithm development as a single falsifiable dry-to-wet loop.

Why it matters: This paper presents a concrete, recent AI system with measurable improvements in both computational benchmarks and prospective wet-lab validations, offering significant value for readers interested in autonomous scientific discovery and enzyme engineering.

arxiv1d agoScore 85
enzyme engineeringrecursive self-improvementautonomous laboratories
Research Editor selected

Automated Discovery of Datasets with Surprising Side Effects on Language Models

Researchers propose a method to automatically discover training datasets that cause unexpected side effects in language models, such as inducing misalignment or outdated personas. They define three objectives for these datasets and use LLM judges with evolutionary and reinforcement learning to find them, producing 6,575 datasets across 8 behaviors. The findings show that automated metrics align with human judgments and that frontier models can predict side effects better than humans.

Why it matters: This paper introduces a novel automated pipeline for discovering training data side effects, offering a testbed for studying generalization and improving model safety.

arxiv1d agoScore 85
language model generalizationtraining data discoveryAI safety
Research Editor selected

Prism: Bidirectional Self-Teaching for Unified Multimodal Models

Prism is a post-training method for unified multimodal models that derives bidirectional supervision from self-generated interleaved rollouts)Skip. It decouples trajectories into text and image chains, assesses their reliability, and constructs training targets accordingly. Experiments on SenseNova-U1 and BAGEL show improvements in both generation and understanding, e.g., GenEval +3.4 and MME +29 points, without external verifiers or annotated corrections.

Why it matters: This presents a novel training method with concrete benchmark gains, offering practical value for multimodal model development.

arxiv1d agoScore 82
Unified Multimodal ModelsSelf-supervised LearningModel Post-training
Research Editor selected

Incentivizing Learning in AI-Assisted Work: A Model of Skill Dynamics

This paper extends an accept-or-solve model to sequential tasks where worker skills evolve through independent practice or reliance on AI. It shows that myopic workers can get trapped at low skill levels or lose skills, while strategic workers may choose independent work to preserve future capabilities. The authors characterize incentive payments that encourage skill acquisition and retention, finding that independent work can become self-sustaining and that higher skill targets may reduce long-run costs.

Why it matters: Provides a concrete economic model and actionable insights for organizations designing AI-assisted workflows to sustain human expertise.

arxiv1d agoScore 85
AI-assisted workskill dynamicsincentive design
Research Editor selected

Traffic-VL: Reinforced Vision-Language Model for End-to-End Traffic Signal Control

Traffic-VL is an 8-billion-parameter vision-language model that directly maps multi-view surveillance images and natural-language phase descriptions to traffic signal decisions, bypassing intermediate structured states. It is trained with offline reinforcement learning on photorealistic synthetic data (SynTra-V) and online stepwise trajectory optimization, matching structured-state baselines while outperforming image-based controllers and general-purpose VLMs. Evaluations include unseen topologies, degraded sensing, real footage, and a three-week deployment across ten intersections.

Why it matters: This paper presents a concrete, recent AI development with strong experimental validation, including real-world deployment, offering practical value for autonomous traffic management.

arxiv1d agoScore 85
Traffic signal controlVision-language modelsReinforcement learning
Research

Corralling AI Risks with Grassroots Contracts

The paper proposes grassroots contracts, a digital adaptation of Proudhon's mutualism, to address three major AI risks on digital platforms: AI control of global platforms, unaccountable AI agents, and unattributed deepfakes. It presents implementations for a social graph with accountable AI agents and a social network with digital provenance, along with working smartphone apps. The approach aims to restore accountability and provenance through decentralized, user-driven agreements.

arxiv1d agoScore 72
AI governancedigital platformsAI agents
Research

Context Compression Weakens Refusal Alignment in LLMs

A new study shows that encoder-based context compression systematically weakens refusal behavior in instruction-tuned language models, even when reconstruction fidelity is high. The effect persists across model families and compression architectures, and is traced to attenuation of the decoder's refusal direction. The authors propose Memory Steering, a lightweight inference-time intervention that restores refusal rates without retraining.

arxiv1d agoScore 78
AI safetymodel compressionalignment
Research

LegalEvolver: Rubric-Guided Self-Evolution for Legal Judgment Prediction

LegalEvolver is a query optimization framework for legal judgment prediction that improves retrieval by generating legally informative queries. It uses heuristic seed queries and rubric-guided reinforcement learning to iteratively self-evolve, surpassing initial performance ceilings. Benchmarks show effectiveness over direct SFT+RL and fixed rubrics.

arxiv1d agoScore 78
Legal Judgment PredictionQuery OptimizationSelf-Evolution
Research Editor selected

Process-based catchment model matches deep-learning accuracy across three continents

A new process-based hydrological model, CaRT+, built on physical laws and refined with AI assistance, achieves accuracy comparable to LSTM neural networks on catchments in Britain, the US, and Australia. With 35 parameters per catchment and no daily neural network, it reaches median Nash-Sutcliffe efficiencies of 0.863 and 0.765 in held-out years on British and US data, and outperforms LSTMs on Australian catchments. The results suggest that catchment-scale laws exist as laws of form, with shared constitutive families and only parameters learned.

Why it matters: This is a concrete, recent research result showing a process-based model matching deep learning accuracy, with implications for hydrology and AI-assisted model refinement.

arxiv1d agoScore 85
hydrological modelingdeep learning comparisonphysical laws
Research Editor selected

MA-Mesh: Role-based Self-Healing for Resilient Multi-Agent Collaboration

MA-Mesh is a framework that adds a role-based self-healing layer to multi-agent systems, generating resilience-aware specifications and executable policy code. It introduces cross-agent reflection and role-based failover to improve robustness under uncertainty. Experiments across software development, tool use, and scientific reasoning show significant gains in performance and robustness.

Why it matters: Provides a concrete, novel framework with explicit resilience mechanisms and experimental validation, valuable for AI system design.

arxiv1d agoScore 82
multi-agent systemsresilience engineeringrole-based coordination
Research

GIST: Gibbs-Targeted Hierarchical Data Selection for Offline RLVR

This paper introduces GIST, a two-level data selection method for offline reinforcement learning with verifiable rewards (RLVR). It derives a factorization of the KL divergence between the current policy and the Gibbs-optimal target, identifying which responses and tokens are most informative. GIST selects correct responses on-policy and incorrect responses near the policy's mode, and supervises only the tokens where the mismatch lies. Experiments show accuracy gains across GSM8K, MATH-500, and MBPP for three models, with 66.6% fewer supervised tokens in a controlled comparison.

arxiv1d agoScore 78
Reinforcement LearningData SelectionToken Efficiency
Research Editor selected

Zero-Shot Coordination among LLM Agents

This paper investigates zero-shot coordination (ZSC) in LLM agents, finding that even frontier models struggle to coordinate in simple, controlled environments due to limited understanding of the coordination problem and weak reasoning about partners' beliefs. The authors propose Coordination-friendly definitions (CFDs) as a principled approach to enable robust coordination)Skip

Why it matters: Provides concrete experimental findings and a novel method (CFDs) for improving LLM coordination, relevant to AI research and deployment.

arxiv1d agoScore 82
LLM coordinationzero-shot coordinationmulti-agent systems