AI.info

Latest Research — Page 2

Browse Latest Research on AI.info.

• Decouple Searching from Training: Scaling Data Mixing via Model Merging for Large Language Model Pre-training
• Transformers Provably Learn Chain-of-Thought Reasoning with Length Generalization
• ReTraceQA: Evaluating Reasoning Traces of Small Language Models in Commonsense Question Answering
• Fast Mining and Dynamic Time-to-Event Prediction over Multi-sensor Data Streams
• NTIRE 2025 Challenge on Low Light Image Enhancement: Methods and Results
• A solvable high-dimensional model where nonlinear autoencoders learn structure invisible to PCA while test loss misaligns with generalization
• From Rollouts to Recipes: Self-Contained Post-Training for LLMs
• Benchmarking Language Modeling for Lossless Compression of Full-Fidelity Audio
• Style Amnesia: Investigating Speaking Style Degradation and Mitigation in Multi-Turn Spoken Language Models
• MEG-to-MEG Transfer Learning and Cross-Task Speech/Silence Detection with Limited Data
• Detecting and Mitigating Memorization in Diffusion Models through Anisotropy of the Log-Probability
• HE-SNR: Uncovering Latent Logic via Entropy for Guiding Mid-Training on SWE-bench
• Learning Interestingness in Automated Mathematical Theory Formation
• UniCon: A Unified System for Efficient Robot Learning Transfers
• MCMoE: Completing Missing Modalities with Mixture of Experts for Incomplete Multimodal Action Quality Assessment
• JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion
• The Stackelberg Speaker: Optimizing Persuasive Communication in Social Deduction Games
• Assessing Automated Fact-Checking for Medical LLM Responses with Knowledge Graphs
• TEAS: Trusted Educational AI Standard: A Framework for Verifiable, Stable, Auditable, and Pedagogically Sound Learning Systems
• Semantic Radiance Fields as Simulators for Spatial Reasoning in Real-World Scenes
• How Does Label Noise Gradient Descent Improve Generalization in the Low SNR Regime?
• Aster: Autonomous Scientific Discovery over 20x Faster Than Existing Methods
• CGCE: Classifier-Guided Concept Erasure in Generative Models
• HarmTrace: Anchor-Calibrated Decoupled Optimization for Fine-Grained Target Identification in Harmful Memes
• Incorporating Self-Rewriting into Large Language Model Reasoning Reinforcement
• FURINA: A Fully Customizable Role-Playing Benchmark via Scalable Multi-Agent Collaboration Pipeline
• HLPD: Aligning LLMs to Human Language Preference for Machine-Revised Text Detection
• Debias-SparseGPT: Bias-Aware Pruning for Large Language Models
• Spherical Steering: Geometry-Aware Activation Rotation for Language Models
• HeteroCache: A Dynamic Retrieval Approach to Heterogeneous KV Cache Compression for Long-Context LLM Inference
• Readability Reconsidered: A Cross-Dataset Analysis of Reference-Free Metrics
• MoD-DPO: Towards Mitigating Cross-modal Hallucinations in Omni LLMs using Modality Decoupled Preference Optimization
• MS-BART: Unified Modeling of Mass Spectra and Molecules for Structure Elucidation
• InertialAR: Autoregressive 3D Molecule Generation with Inertial Frames
• Computational Foundations for Strategic Coopetition: Formalizing Trust and Reputation Dynamics
• BAID: A Benchmark for Bias Assessment of AI Detectors
• Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents
• Auto-Regressive Masked Diffusion Models
• MixRI: Mixing Features of Reference Images for Novel Object Pose Estimation
• Seg-VAR: Image Segmentation with Visual Autoregressive Modeling