How I Study AI - Learn AI Papers & Lectures the Easy Way

Efficient RLVR Training via Weighted Mutual Information Data Selection

Intermediate

Xinyu Zhou, Boyu Zhu et al.Mar 2arXiv

Reinforcement learning (RL) trains language models by letting them try answers and learn from rewards, but training is slow if we pick the wrong practice questions.

#Reinforcement Learning#RLVR#Data Selection

Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs

Intermediate

Yu Liang, Zhongjin Zhang et al.Feb 2arXiv

This paper proposes ReSID, a new way to turn items into short token codes (Semantic IDs) that are much easier for a recommender to predict.

#Semantic IDs#Generative Recommendation#Representation Learning

Schoenfeld's Anatomy of Mathematical Reasoning by Language Models

Intermediate

Ming Li, Chenrui Fan et al.Dec 23arXiv

This paper turns messy chains of thought from language models into clear, named steps so we can see how they really think through math problems.

#Schoenfeld’s Episode Theory#Cognitive Episodes#ThinkARM

Papers3

Efficient RLVR Training via Weighted Mutual Information Data Selection

Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs

Schoenfeld's Anatomy of Mathematical Reasoning by Language Models