Translate
Live intelligence

Latest AI updates

1,700 updates
AllModelsProductsIndustryResearchTutorialOpinion
Research Editor selected

ModalSearcher: Autonomous Cross-Modal Search Agent for Multimodal QA

ModalSearcher introduces autonomous cross-modal search, enabling agents to dynamically choose between reasoning, text search, image search, and reverse-image search. The authors propose a pipeline to build interleaved multimodal knowledge graphs and train an 8B agent with SFT and RL, achieving 16.2-point average accuracy gains over Qwen3-VL-8B on five benchmarks.

Why it matters: This paper presents a concrete method and benchmark for improving multimodal agents' ability to handle knowledge-intensive tasks, with measurable gains.

arxiv1d agoScore 82
multimodal large language modelsautonomous searchknowledge-intensive QA
Research Editor selected

SafeStart: Mitigating the Overlooked Safety Cost of Active Expert Reduction in MoE LLM Serving

This paper reveals that expert-reduction methods for MoE LLM serving, which speed up inference by running fewer experts per token, can weaken refusal of harmful requests. The authors introduce SafeStart, which runs all experts for the prompt and first eight generated tokens, then reduces experts afterward, preserving most speedup while restoring safety. In vLLM on an RTX A6000, SafeStart keeps at least 85% of the offline decode speedup at batch sizes 8 and 32 and reduces harmful compliance increases to near baseline levels across three models.

Why it matters: Provides a concrete, recent finding that expert reduction in MoE serving has safety costs and offers a practical mitigation, valuable for AI serving systems and safety research.

arxiv1d agoScore 85
Mixture-of-ExpertsLLM safetyInference optimization
Research

Text-Only Training for Zero-Shot Video Temporal Localization

This paper explores training a text-only LLM on synthetic textual video descriptions to perform temporal grounding, receiving video only at inference. The authors generate a 134K-example corpus, adapt Qwen3-8B with LoRA on text embeddings, and propose Nearest Entity Replacement (NER) to bridge the modality gap. NER improves zero-shot performance on Charades-STA and QV-Highlights, but a significant gap remains.

arxiv1d agoScore 78
Video Temporal LocalizationMultimodal LearningZero-Shot Transfer
Research Editor selected

CARM: Conformal Adaptive Risk Masking for Reliable Scientific Super-Resolution

CARM is a statistical framework that adapts conformal risk control to temporal distribution shifts in Earth-observation super-resolution. It aggregates threshold proposals from experts with different update rates using an asymmetric surrogate loss, enabling rapid adaptation while maintaining stability. Theoretical guarantees show time-averaged mask-conditioned risk converges to at most alpha at O(1/T) rate. On South China Sea chlorophyll-a data, CARM reduces empirical violation rate from 43.49% to 9.38% at alpha=0.10, with consistent improvements on ERA5 and OSTIA datasets.

Why it matters: Provides a novel, theoretically grounded method for reliable AI in scientific super-resolution, with strong empirical results across multiple datasets.

arxiv1d agoScore 82
Conformal risk controlAdaptive thresholdingScientific super-resolution
Research Editor selected

Reasoning-Aware Cross-Modal Association Unlearning in Multimodal Large Reasoning Models

This paper introduces cross-modal association unlearning for multimodal large reasoning models (MLRMs), targeting the removal of specific visual-to-knowledge associations rather than whole samples or concepts. The proposed VKCR method localizes edits to a compact parameter subspace and counterfactually rebinds reasoning behavior to suppress target leakage while preserving non-sensitive knowledge. New benchmarks CLEAR-R and MLLMU-R evaluate leakage, hallucination, and utility, showing VKCR reduces image-conditioned target leakage across tasks.

Why it matters: Addresses a novel privacy problem in multimodal reasoning models with a parameter-efficient method and benchmarks, offering practical value for AI safety and unlearning research.

arxiv1d agoScore 82
machine unlearningmultimodal large reasoning modelscross-modal association
Research Editor selected

AssameseLM-E2B: An Open-Weight Assamese Language Model for Efficient Consumer-Grade Deployment

This paper introduces AssameseLM-E2B, a 2-billion-parameter language model adapted for the low-resource Assamese language, built on the Gemma-4 architecture with an expanded byte-level BPE vocabulary. It uses parameter-efficient continual pre-training with Hybrid LoRA and supervised fine-tuning to improve instruction-following and RAG capabilities. The model achieves a BPB of 0.7488 on FLORES-200, a RAG score of 96.7%, and 100% refusal accuracy on adversarial unanswerable prompts, while highlighting limitations in general-purpose baselines.

Why it matters: Provides a concrete, recent advancement in low-resource language modeling with practical deployment implications, offering valuable insights for developers and researchers.

arxiv1d agoScore 85
Low-resource language modelsParameter-efficient fine-tuningRetrieval-augmented generation
Research Editor selected

GEdit: Geometry-Aware Regularization for Knowledge Editing in LLMs

GEdit is a new framework for knowledge editing in large language models that uses a Low-Rank Plus Diagonal (LR+D) factor model to capture the geometric structure of Transformer FFN hidden representations. It introduces low-rank regularization to protect core knowledge and adaptive regularization that adjusts strength based on edit difficulty. Experiments on GPT2-XL, GPT-J, and LLaMA-3-8B with up to 10,000 continuous updates show consistent improvements over baselines in both editing success and knowledge preservation.

Why it matters: This paper presents a novel, technically grounded approach to a practical LLM challenge (knowledge editing) with strong empirical results across multiple models and large-scale settings.

arxiv1d agoScore 82
knowledge editinglarge language modelsregularization
Research Editor selected

CoSafe: Cooperative Runtime Safety Assurance for Mixed-Traffic Autonomous Driving

CoSafe is a modular, ADS-agnostic framework that extends runtime safety assurance from ego-centric takeover to scene-level cooperative intervention in mixed-traffic autonomous driving. It uses V2X communication and roadside units to monitor hazards and jointly mitigate risks involving multiple vehicles. Integrated with seven state-of-the-art ADSes, CoSafe improves driving scores by 70.03% on average, reduces safety violations by 87.80%, and achieves an F3 score of 0.930 for unsafe AV identification while meeting real-time constraints.

Why it matters: This paper presents a concrete, recent advancement in autonomous driving safety with measurable improvements and a novel cooperative approach, offering practical value for researchers and developers.

arxiv1d agoScore 82
autonomous driving safetycooperative systemsruntime monitoring
Research

A Definition of Catastrophic Performance

This paper introduces a formal definition of catastrophic performance to distinguish suboptimal AI outcomes from truly catastrophic ones. The definition relies on two reference values: baseline value (no agent deployed) and orthogonal value (a trajectory unrelated to human preferences). Catastrophic performance is characterized as an agent undoing benefits from past human activity, enabling rigorous analysis of AI misalignment risk.

arxiv1d agoScore 72
AI alignmentcatastrophic riskvalue specification
Research Editor selected

Rubric Curriculum RL Improves RLVR in Non-Verifiable Domains

The paper introduces Rubric Curriculum RL (RcRL), a method that exploits the generation-verification gap in domains without ground-truth rewards by combining pairwise preferences over rollouts with a curriculum over rubric criteria. This creates a more discriminative and non-stationary training signal, avoiding the plateau seen in baselines. RcRL achieves a 70.5% human win rate on creative writing and improves HealthBench performance by 14.6% (full set) and 25.4% (consensus subset), outperforming all training baselines.

Why it matters: This presents a concrete, novel training method with measurable gains in open-ended domains, addressing a key limitation of RLVR.

arxiv1d agoScore 82
Reinforcement LearningReward ModelingLLM Alignment
Research Editor selected

HarnessFlow: MCTS-Guided Harness Optimization for LLM Agents

HarnessFlow is a framework that automatically optimizes the harness (the code governing tool interaction) for LLM agents under a limited evaluation budget. It uses a Monte Carlo Tree Search policy to balance exploration and exploitation, with a coding agent refining harnesses based on execution feedback. Across six benchmarks, it achieves an average score of 69.48%, outperforming initial harnesses by 10.89 points and Meta-Harness by 3.95 points, while reducing per-task execution costs.

Why it matters: This introduces a novel, practical method for automating harness design, showing significant performance gains and cost efficiency, which is valuable for AI agent development.

arxiv1d agoScore 85
LLM agentsharness optimizationMCTS
Research Editor selected

MPRBench: A New Benchmark for Multimodal Personalized Reasoning in Vision-Language Models

MPRBench is a comprehensive benchmark introduced to evaluate multimodal personalized reasoning in large language models, covering 12 sub-tasks and supporting interpretable error analysis. Experiments show current MLLMs struggle with recognizing personalized concepts and are sensitive to irrelevant personal information. The benchmark aims to drive further research in this area.

Why it matters: Provides a concrete, recent benchmark with findings that highlight current limitations in MLLM personalized reasoning, valuable for researchers and developers.

arxiv1d agoScore 80
Multimodal Large Language ModelsPersonalized ReasoningBenchmarking
Research Editor selected

Pseudo-Riemannian Attention Network (PRAT) for Heterophilic Graphs

This paper identifies a geometric mismatch in Graph Neural Networks on heterophilic graphs, where positive-definite metrics fail to capture dissimilarity. The authors introduce the Graph Signature Index to measure this mismatch and propose PRAT, a network using a learnable pseudo-Riemannian metric. PRAT outperforms or matches strong baselines on fifteen node-classification benchmarks, with ablations confirming the importance of the indefinite metric.

Why it matters: This work offers a novel theoretical insight and a practical method that improves GNN performance on heterophilic graphs, a challenging and active research area.

arxiv1d agoScore 82
Graph Neural NetworksGeometric Deep LearningHeterophily
Research Editor selected

NextGAN: Towards Simple and Scalable GAN Training for Image and Video Generation

NEXTGAN trains a scalar critic by regressing its input gradient onto displacements between independently paired reference and generated features at interpolated locations in frozen representation spaces. The authors state that, for a fixed generator, this regression is equivalent to an adversarial objective with a gradient-energy penalty, while the generator keeps the standard potential-based GAN update. The method adapts pre-trained image and video generators with up to 14B parameters without auxiliary losses, online teacher scores, or trajectories. On ImageNet 256x256 it reports one-step FIDs of 2.87 with pixel-space JiT-B and 1.54 with latent-space DiT-XL/1 (1.36 for its adversarial form), and a single Wan2.1-1.3B model supports one-, two- and four-step text-to-video sampling, scoring 83.13 VBench total at two steps.

Why it matters: It offers a simpler adversarial training recipe for few-step image and video generation, backed by one-step FID and VBench comparisons, which is useful for researchers and practitioners building fast generative models.

arxiv1d agoScore 80
NextGANgenerative adversarial networkstext-to-video generation
Research

OccSplat Targets Occluded Geometry in Driving-Scene Gaussian Splatting

Researchers propose OccSplat, a feed-forward 3D Gaussian Splatting framework for reconstructing driving scenes from surround-view images. Rather than inpainting missing pixels in rendered views, the method identifies regions lacking Gaussian support via Gaussian Occlusion Mapping and generates new primitives through Gaussian Occlusion Completion. These primitives are merged into the original representation, keeping feed-forward inference without per-scene optimization. Experiments on nuScenes show improvements over compared methods, with the largest gains reported in occluded regions.

arxiv1d agoScore 72
3D Gaussian Splattingautonomous drivingnovel-view synthesis
Research Editor selected

DyPSI: Joint Generation of Physics Fields and Sensor Trajectories for Dynamic Sensing

DyPSI is a new framework for dynamic physics sensing that jointly generates future spatiotemporal fields and sensor trajectories from historical observations. It uses continuous Sensor Position Fields, a shared Functional Tucker representation, and a joint diffusion model to produce temporally coherent sensing plans. Experiments on turbulent flow, sea-surface temperature, and car aerodynamics show improved error reduction over static and frame-wise placement strategies.

Why it matters: This paper introduces a novel method that addresses a key limitation in sensor placement by jointly optimizing field reconstruction and trajectory generation, with demonstrated gains on multiple real-world datasets.

arxiv1d agoScore 82
sensor placementdiffusion modelsphysics-informed AI
Research

MonoLaDN: Language-Guided DeNoising for Monocular 3D Object Detection

MonoLaDN is a training framework that transfers geometric reasoning from a frozen vision-language model into a monocular 3D detector. It encodes the VLM's structured reasoning about occlusion, depth, dimension, and angle into queries, perturbs them with noise, and trains the detector to reconstruct the clean reasoning. The deployed detector remains purely vision-based, adding no inference latency or parameters. Experiments on KITTI 3D and nuScenes report consistent improvements over state-of-the-art detectors and good generalization to new domains.

arxiv1d agoScore 78
monocular 3D object detectionvision-language modelsknowledge distillation
Research

The Split Protocol Decides the Result: Re-Evaluating Encrypted Traffic Classification

This arXiv paper shows that encrypted traffic classifiers can score macro-F1 0.992 under a standard evaluation protocol but only 0.567 when flows from one packet capture are kept on one side of the split. Across five split protocols on backbone QUIC traffic, server-disjoint splitting costs 0.192 macro-F1, while random, temporal and client-disjoint splits are statistically indistinguishable from session-disjoint splitting. The authors propose single-identifier probes as a pre-training diagnostic for identifying the relevant split axis, and report that several pre-registered architectural claims fail under the corrected evaluation.

arxiv1d agoScore 76
encrypted traffic classificationevaluation methodologydata leakage