Explore Broadly, Reason Sharply: Push Small Models toward the Frontier via Sampling
Paper • 2609.38104 • Published • 3
Joakim Lee
Reinforcement4All
AI & ML interests
None yet
Recent Activity
upvoted a paper about 16 hours ago
Explore Broadly, Reason Sharply: Push Small Models toward the Frontier via Sampling upvoted a paper about 17 hours ago
Before It Fades: Reinforcing Temporal Representations at Inference Time in VideoLLMs upvoted a paper about 17 hours ago
JevSpawn: Adaptive Agentic Inference through Compositional Action SpacesOrganizations
None yet
upvoted a paper about 16 hours ago
upvoted 19
papers about 17 hours ago
Before It Fades: Reinforcing Temporal Representations at Inference Time in VideoLLMs
Paper • 2610.01595 • Published • 2
JevSpawn: Adaptive Agentic Inference through Compositional Action Spaces
Paper • 2610.00437 • Published • 3
Joint and Cross-Modal Video-Audio Generation and Editing: A Unified Formulation and Design Taxonomy
Paper • 2609.34381 • Published • 3
Personalized Image Generation with Reasoning and Reflection
Paper • 2610.00737 • Published • 4
VTR-Bench: A Systematic Benchmark for Evaluating Visual Text Rendering in Video Generation
Paper • 2610.01499 • Published • 7
ScholarCatalyst: A Benchmark for Retrieving Papers That Inspire New Research
Paper • 2610.02202 • Published • 6
Memorizon: Training World Models Beyond Their Context Window
Paper • 2610.00544 • Published • 5
LOCI: Spatial Linear Memory for Streaming World Models
Paper • 2609.40222 • Published • 6
SemanTok: Predictable Semantic Tokens for Efficient Autoregressive Video Generation
Paper • 2610.00686 • Published • 7
RLE-Bench: A Qualifying Exam for Coding Agents as Robot Learning Engineers
Paper • 2609.34210 • Published • 6
OmniSeek: Native Tool Integration for Multi-turn Audio-Visual Reasoning
Paper • 2610.02181 • Published • 8
Better Supervision Is Nearby: Neighborhood On-Policy Self-Distillation
Paper • 2609.39687 • Published • 10
Pretrain Once, Route Anywhere: Towards a Foundation Model for LLM Routing
Paper • 2609.37362 • Published • 9
Benchmarking and Enhancing Skill-Level Memory for Partially Observable Robotic Manipulation
Paper • 2609.38886 • Published • 8
PixelDense: Dense Prediction as Representation Alignment for Pixel Diffusion
Paper • 2610.00483 • Published • 13
Decoding Looped Transformers Better for (Almost) Free
Paper • 2610.02185 • Published • 19
Omni-Embed-Mini: Binding Modalities Without Forgetting via Dense Distillation
Paper • 2610.02148 • Published • 15
CorrGRPO: Correlation-Normalized GRPO for Multi-Reward Learning
Paper • 2609.36820 • Published • 18
SILSA: Sliding-Window Slice Latents for Topology-Preserving High-Resolution 3D Generation
Paper • 2610.02201 • Published • 19