FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders
Paper • 2609.31620 • Published • 158
Finn Reyes
finnreyes
AI & ML interests
Retrieval-augmented generation, dense retrieval, open-domain question answering, RAG pipelines
Recent Activity
liked a model 2 days ago
Neesham/semantic-searchOrganizations
None yet
upvoted a paper 1 day ago
upvoted 6
papers 2 days ago
Paragraph Boundaries Are Not White Space:Compression Depth as the Signature of Hierarchical Structure
Paper • 2609.23551 • Published • 7
D-JEPA: A Decision-Aligned Latent World Model
Paper • 2609.24749 • Published • 17
Disaggregated Quantization: Specializing LLM Prefill and Decode
Paper • 2609.26333 • Published • 91
TimeEvo: Failure-Driven Self-Evolution of a Time Series Agent
Paper • 2609.27277 • Published • 31
Not All Ranks Are Equal: Budget-Aware LoRA Merging Across Tasks
Paper • 2609.22237 • Published • 7
Game Arena: Strategic LLM Evaluation in Competitive Environments
Paper • 2609.31473 • Published • 12
upvoted 6
papers 3 days ago
RGBD20K: A Large-Scale Benchmark for RGB-D Semantic Segmentation
Paper • 2609.29028 • Published • 15
DeltaWAM: Delta World Action Models for Bimanual Manipulation
Paper • 2609.28811 • Published • 18
World Action Agent: Harnessing VLMs for Robot Manipulation via World Action Rehearsal
Paper • 2609.29964 • Published • 15
Agent-Editing World Model: Rethinking World Modeling for LLM Agents
Paper • 2609.28416 • Published • 42
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs
Paper • 2609.29845 • Published • 100
Coding Agents for Generalized Task and Motion Planning Problems
Paper • 2609.30233 • Published • 24
upvoted 4
papers 4 days ago
HARMONY: Hierarchical Agentic Reasoning for MONocular Image-to-Scene Synthesis
Paper • 2609.26793 • Published • 9
LatentPort: Beyond KV Cache - Cross-Model Transfer of Recurrent Memory in Hybrid Language Models: A 4B-to-9B Hybrid-State Handoff Without Target Prefix Replay
Paper • 2609.25053 • Published • 17
RoboFollow: Unveiling the Instruction Following Mirage in Embodied Agents
Paper • 2609.25636 • Published • 13
RULER: Instance-aware Rubric Rewards for SVG Generation
Paper • 2609.25270 • Published • 102
upvoted 3
papers 5 days ago
GraphSkillEvo: Evolutionary Optimization of Graph-Structured Agent Skills
Paper • 2609.21749 • Published • 19
The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction
Paper • 2609.18063 • Published • 19
CADWorld: Computer-Use Benchmark for Long-Horizon Computer-Aided Design
Paper • 2609.16251 • Published • 14