AI.info
Latest Research — Page 24
Browse Latest Research on AI.info.
- AVATAAR: Agentic Video Answering via Temporal Adaptive Alignment and Reasoning
- Deep SPI: Safe Policy Improvement via World Models
- EdgeDeID: A 0.6B Clinical De-identifier Trained Without Real Patient Data
- BridgeTwist: Twisting Schrödinger Bridges for Training-Free Conditional Sampling
- StdGEN++: A Comprehensive System for Semantic-Decomposed 3D Character Generation
- FlowBind: Efficient Any-to-Any Generation with Bidirectional Flows
- Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents
- FreeAskWorld: An Interactive and Closed-Loop Simulator for Human-Centric Embodied AI
- Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design
- ScholarEval: Research Idea Evaluation Grounded in Literature
- VAGEN: Reinforcing World Model Reasoning for Multi-Turn VLM Agents
- Private Zeroth-Order Optimization with Public Data
- Learning Reconfigurable Representations for Multimodal Federated Learning with Missing Data
- EagleVision: A Dual-Stage Framework with BEV-grounding-based Chain-of-Thought for Spatial Intelligence
- TOSC: Task-Oriented Shape Completion for Open-World Dexterous Grasp Generation from Partial Point Clouds
- Perceive, Act and Correct: Confidence Is Not Enough for Hyperspectral Classification
- Why Do Video Diffusion Models Violate Physics? Unveiling the Flaws in Attention Mechanisms
- Zero-Shot Instruction Following in RL via Structured LTL Representations
- OnlineSplatter: Pose-Free Online 3D Reconstruction for Free-Moving Objects
- Ensuring Calibration Robustness in Split Conformal Prediction Under Adversarial Attacks
- VOYAGER: A Training Free Approach for Generating Diverse Datasets using LLMs
- EgoXperience: Learning World-Grounded Egocentric Representations from Human Experience
- All Vehicles Can Lie: Efficient Adversarial Defense in Fully Untrusted-Vehicle Collaborative Perception via Pseudo-Random Bayesian Inference
- Predicting the Future by Retrieving the Past
- Self-Evolving Coding Agents via Harness-Policy Optimization
- Get RICH or Die Scaling: Profitably Trading Inference Compute for Robustness
- Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight
- FaithLens: Detecting and Explaining Faithfulness Hallucination
- Do We Really Need Multimodal Emotion Language Models Larger Than 1B Parameters?
- Are LLM Evaluators Really Narcissists? Sanity Checking Self-Preference Evaluations
- Virtually Being: Customizing Camera-Controllable Video Diffusion Models with Multi-View Performance Captures
- SkillSeek: Revisiting Agent Skill Retrieval at Marketplace Scale
- Unmasking Reasoning Processes: A Process-aware Benchmark for Evaluating Structural Mathematical Reasoning in LLMs
- PatchAlign3D: Local Feature Alignment for Dense 3D Shape Understanding
- Beyond ReAct: A Planner-Centric Framework for Complex Tool-Augmented LLM Reasoning
- No One to Blame: A Framework of Constitutive AI Unaccountability
- Mapping Faithful Reasoning in Language Models
- ProCARE: Real-World Study Automation via Profile-Grounded Evidence Contracts
- Structure is Supervision: Multiview Masked Autoencoders for Radiology
- Weight-Space Mixture-of-Experts for Implicit Neural Representation Classification