UAD: Unsupervised Affordance Distillation for Generalization in Robotic Manipulation
Paper: https://
arxiv.org/pdf/2506.09284
v1.pdf
…
Page: https://
unsup-affordance.github.io
Code: https://
github.com/TangYihe/unsup
-affordance
… (soon)
@jiqizhixin
-
Unsupervised Affordance Distillation for Robotic Manipulation Generalization
By
–
-
UAD: Unsupervised Affordance Distillation for Robot Learning
By
–
UAD (Unsupervised Affordance Distillation) — a powerful new pipeline for learning fine-grained object affordances without any manual labels. It enables robots to understand how to interact with novel objects in open-ended, real-world scenarios. UAD distills implicit knowledge
-
Offline Domain Randomization: Formal Theory for Sim-to-Real Transfer
By
–
Bridging Sim-to-Real with Theory: Offline Domain Randomization Gets Formal This work strengthens the theoretical underpinnings of Offline Domain Randomization (ODR) — a strategy that uses real-world data to inform simulation training, offering a smarter alternative to
-
DeepSeek’s nano-vllm: lightweight Python vLLM implementation
By
–
DeepSeek engineer Xingkai Yu just dropped nano-vllm — a lightweight vLLM reimplementation from scratch. It's a clean implementation under 1,200 lines of Python code with comparable inference speeds to vLLM. Perfect for learning or building lean! Project:
-
Optimal Advantage Regression Accelerates RL for LLM Reasoning
By
–
Accelerating RL for LLM Reasoning with Optimal Advantage Regression
Paper: https://
arxiv.org/pdf/2505.20686
.pdf
…
Code: https://
github.com/ZhaolinGao/A-PO -
A*-PO: Efficient Two-Stage Policy Optimization for LLMs
By
–
Current RL fine-tuning for LLMs suffers from expensive training, critic networks or advantage estimates, and inefficient exploration. Enter A*-PO: A two-stage policy optimization method that flips the standard approach — slashing compute and memory costs compared to
-
Diffuse and Disperse: Image Generation with Representation Regularization
By
–
Diffuse and Disperse: Image Generation with Representation Regularization
Paper: https://
arxiv.org/pdf/2506.09027
v1.pdf
… -
Diffusion Models: Improving Internal Representations Through Explicit Regularization
By
–
Diffusion models have been wildly successful for image and text generation, but they rely on regression-based objectives and lack explicit representation regularization. As a result, their internal representations are underutilized or poorly structured. This paper introduces
-
IBM Breakthrough in Quantum Error Correction Targets 2029
By
–
IBM Says It’s Cracked Quantum Error Correction A new chip design and architecture has IBM aiming for fault tolerance by 2029
-
Cartridges: Lightweight Long Context Representations via Self-Study
By
–
Cartridges: Lightweight and general-purpose long context representations via self-study
Paper: https://
arxiv.org/pdf/2506.06266
.pdf
…
Code: https://
github.com/HazyResearch/c
artridges
…
And also a wonderful blog: https://
hazyresearch.stanford.edu/blog/2025-06-0
8-cartridges
…