Lattice AI Research

Research focus

Reasoning & Chain-of-Thought (2)RLHF & Preference Learning (1)Training Efficiency & Optimization (1)Data Curation & Synthetic Data (1)Multimodal Models (1)

Frequent co-authors

Xitai Jiang (1)Wenze Lin (1)Yang Yue (1)Gao Huang (1)

Papers (2)

May 21, 2026

Tsinghua AIMay 21, 2026·also Aether Tech, PKU

From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning

Decomposing complex reasoning problems into verifiable subproblems unlocks significant performance gains in LLM reasoning, especially on hard problems previously stuck in gradient dead zones.

Xitai Jiang, Wenze Lin, Yang Yue +2

Reasoning & Chain-of-Thought RLHF & Preference Learning Training Efficiency & Optimization

Mar 17, 2026

Tsinghua AIMar 17, 2026·also DAMO

HopChain: Multi-Hop Data Synthesis for Generalizable Vision-Language Reasoning

Multi-hop data synthesis using HopChain boosts VLM performance across a wide range of tasks, with gains of over 50 points in accuracy for ultra-long-context reasoning.

Shenzhi Wang, Shixuan Liu, Chang Gao +5

Data Curation & Synthetic Data Multimodal Models Reasoning & Chain-of-Thought

Search

Shenzhi Wang

Research focus

Frequent co-authors

Papers (2)