AI.info
Latest Research — Page 44
Browse Latest Research on AI.info.
- Exploring Landscapes for Better Minima along Valleys
- HERMES: Towards Efficient and Verifiable Mathematical Reasoning in LLMs
- Lost in Serialization: Invariance and Generalization of LLM Graph Reasoners
- ACoT-VLA: Action Chain-of-Thought for Vision-Language-Action Models
- Exploring The Empirical Privacy Landscape of Synthetic Data
- Weaving Visual Narratives: Agentic Image Bundle Composition Beyond Atomic Visual Matching
- A Conditional No-Hair Theorem for Transformers: How Language Models Forget Their Prompt
- Point Ladder Tuning: Parameter-Efficient Hierarchical Adaptation for 3D Point Cloud Understanding
- UniRestorer++: Scalable Universal Image Restoration with Multi-Granularity Degradation Modeling
- VASTU: Language Models Struggle to Recognize Online Community Values
- Anisotropic Representations Improve Planning in JEPA World Models
- Learning Query-Aware Budget-Tier Routing for Runtime Agent Memory
- Can AI-Generated Persuasion Be Detected? Persuaficial Benchmark and AI vs. Human Linguistic Differences
- CADMorph: Geometry-Driven Parametric CAD Editing via a Plan-Generate-Verify Loop
- ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents
- Classifier Reconstruction Through Counterfactual-Aware Wasserstein Prototypes
- On the Sample Complexity of Differentially Private Policy Optimization
- TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows
- RouteWeaver: Weaving Mode Selection and Execution into Unified LLM Routing
- Lost in the Prompt Order: Revealing the Limitations of Causal Attention in Language Models
- Bootstrapping Physics-Grounded Video Generation through VLM-Guided Iterative Self-Refinement
- From RL Papers to Framework Code: Inclusion and Review in Seven Open-Source Training Frameworks
- HY-Motion 1.0: Scaling Flow Matching Models for Text-To-Motion Generation
- CMuon: Accelerating and Stabilizing Diffusion Transformer Training via Chunked Momentum Orthogonalization
- ABLE: Using Adversarial Pairs to Construct Local Models for Explaining Model Predictions
- Cooperative Pseudo Labeling for Unsupervised Federated Classification
- AlphaBet: Reinforcement Learning for LLM Event Forecasting Agents
- MirrorBench: A Benchmark to Evaluate Conversational User-Proxy Agents for Human-Likeness
- Dissecting Multimodal In-Context Learning: Modality Asymmetries and Circuit Dynamics in modern Transformers
- No Mean Feat: Simple, Strong Baselines for Context Compression
- Protein Counterfactuals via Diffusion-Guided Latent Optimization
- CollectiveKV: Decoupling and Sharing Collaborative Information in Sequential Recommendation
- Decentralized Attention Fails Centralized Signals: Rethinking Transformers for Medical Time Series
- LatentRiskBench: Measuring State-Dependent Risks in Enterprise Agents
- Adversarial Training for Pixel Diffusion
- Spatial4D-Bench: A Versatile 4D Spatial Intelligence Benchmark
- DOS: Distilling Observable Softmaps of Zipfian Prototypes for Self-Supervised Point Representation
- UnHype: CLIP-Guided Hypernetworks for Dynamic LoRA Unlearning
- Beyond Pixels: Benchmarking and Reward-Based Assessing Framework for Visual Spatial Aesthetics
- Mechanistic Evidence for Faithfulness Decay in Chain-of-Thought Reasoning