AI.info

Latest Research — Page 31

Browse Latest Research on AI.info.

• Associative Syntax and Maximal Repetitions reveal context-dependent complexity in fruit bat communication
• world2vec: Mechanistic world models through latent dynamics learning
• Hidden State Poisoning Attacks against Mamba-based Language Models
• ID-Crafter: VLM-Grounded Online RL for Compositional Multi-Subject Video Generation
• A Framework for Low-Effort Training Data Generation for Urban Semantic Segmentation
• Conditional PED-ANOVA: Hyperparameter Importance in Hierarchical & Dynamic Search Spaces
• When Robots Obey the Patch: Universal Transferable Patch Attacks on Vision-Language-Action Models
• When One Moment Isn't Enough: Multi-Moment Retrieval with Cross-Moment Interactions
• Leave It to the Experts: Detecting Knowledge Distillation via MoE Expert Signatures
• ReGAIN: Retrieval-Grounded AI Framework for Network Traffic Analysis
• Source Preference in the Wild: How LLM Agents Favor Items by Source, and How to Reduce It
• Controllable LLM Reasoning via Sparse Autoencoder-Based Steering
• Are we Merging the Right Models? Impact of Expert Training Duration on Model Merging for LLMs
• gfnx: Fast and Scalable Library for Generative Flow Networks in JAX
• Flow Equivariant World Models: Memory for Partially Observed Dynamic Environments
• FineSplat: Fine Patches, Local Groups, Detailed Splats
• Can We Predict Before Executing Machine Learning Agents?
• One Inverse Step is a Convex Program: Bayes-Limit Calibration of Diffusion Inversion
• WebATLAS: An LLM Agent with Experience-Driven Memory and Action Simulation
• Rethinking the Use of Vision Transformers for AI-Generated Image Detection
• Tracing the Latent Threads: A Mechanistic Study of How LLMs Represent and Operationalize Race and Ethnicity Cues
• BanglaSentNet: An Explainable Hybrid Deep Learning Framework for Multi-Aspect Sentiment Analysis with Cross-Domain Transfer Learning
• LEASH: Uncertainty-Aware Weak-to-Strong Escalation and Oversight for Computer-Use Agents
• MeViS: A Multi-Modal Dataset for Referring Motion Expression Video Segmentation
• PRoH: Dynamic Planning and Reasoning over Knowledge Hypergraphs for Retrieval-Augmented Generation
• Diffusion Reward Models
• TalkVerse: Democratizing Minute-Long Audio-Driven Video Generation
• Rethinking Multimodal Few-Shot 3D Point Cloud Segmentation: From Fused Refinement to Decoupled Arbitration
• Thought without systematicity? Evaluating reasoning models on rule induction tasks
• Point-Focused Attention Meets Context-Scan State Space: Robust Biological Visual Perception for Point Cloud Representation
• Structure-R1: Dynamically Leveraging Structural Knowledge in LLM Reasoning through Reinforcement Learning
• Unsupervised Tool Evolution for Coding Agents in Data-Lake Analytics
• TABX: A High-Throughput Sandbox Battle Simulator for Multi-Agent Reinforcement Learning
• JoVA: Unified Multimodal Learning for Joint Video-Audio Generation and Editing
• What Matters for NVFP4 Training? A Comparative Study of Low-Precision Pretraining Recipes
• LT-Soups: Bridging Head and Tail Classes via Subsampled Model Soups
• Bagpiper: Solving Open-Ended Audio Tasks via Rich Captions
• TEXTRIX: Latent Attribute Grid for Native Texture Generation and Beyond
• DeFT-RAG: Incrementally Decomposing Multi-Hop Questions into Self-Contained Sub-Questions with Reusable Answers
• FUME: Fused Unified Multi-Gas Emission Network for Livestock Rumen Acidosis Detection