Daily Papers of 2026-07-06

1. The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning 161 upvotes, #1 of 2026-07-06
2. Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots 56 upvotes, #2 of 2026-07-06
3. DataComp-VLM: Improved Open Datasets for Vision-Language Models 51 upvotes, #3 of 2026-07-06
4. OrbitQuant: Data-Agnostic Quantization for Image and Video Diffusion Transformers 39 upvotes, #4 of 2026-07-06
5. VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon 30 upvotes, #5 of 2026-07-06
6. Measuring the Gap Between Human and LLM Research Ideas 17 upvotes, #6 of 2026-07-06
7. Securing the AI Agent: A Unified Framework for Multi-Layer Agent Red Teaming 14 upvotes, #7 of 2026-07-06
8. MultAttnAttrib: Training-Free Multimodal Attribution in Long Document Question Answering 13 upvotes, #8 of 2026-07-06
9. AnyBokeh: Physics-Guided Any-to-Any Bokeh Editing with Optical Fingerprint Transfer 10 upvotes, #9 of 2026-07-06
10. Teaching LLMs to Recommend and Defer in Underrepresented Epilepsy Care 9 upvotes, #10 of 2026-07-06
11. Generated Contents Enrichment 7 upvotes, #11 of 2026-07-06
12. AGE: Adaptive-masking for Graph Embedding in Graph Retrieval-Augmented Generation 6 upvotes, #12 of 2026-07-06
13. Interpretation-Oriented Cloud Removal via Observation-Anchored Residual Flow with Geo-Contextual Alignment 6 upvotes, #12 of 2026-07-06

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.