Tsinghua-LeapLab

university

https://www.leaplab.ai/

LeapLabTHU

LeapLabTHU

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

andrewzh authored a paper 10 days ago

Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning

andrewzh authored a paper 10 days ago

ExCyTIn-Bench: Evaluating LLM agents on Cyber Threat Investigation

andrewzh authored a paper 10 days ago

Are My Optimized Prompts Compromised? Exploring Vulnerabilities of LLM-based Optimizers

View all activity

Papers

The Flexibility Trap: Why Arbitrary Order Limits Reasoning Potential in Diffusion Language Models

View all Papers

models 0

None public yet

datasets 0

None public yet

Tsinghua-LeapLab (Tsinghua-LeapLab)

Tsinghua-LeapLab

university

https://www.leaplab.ai/

LeapLabTHU

LeapLabTHU

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

andrewzh authored a paper 10 days ago

Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning

andrewzh authored a paper 10 days ago

ExCyTIn-Bench: Evaluating LLM agents on Cyber Threat Investigation

andrewzh authored a paper 10 days ago

Are My Optimized Prompts Compromised? Exploring Vulnerabilities of LLM-based Optimizers

View all activity

Papers

The Flexibility Trap: Why Arbitrary Order Limits Reasoning Potential in Diffusion Language Models

View all Papers

models 0

None public yet

datasets 0

None public yet