Deep Data Research Benchmark
weiliu
thinkwee
AI & ML interests
LLM reasoning, agents
Recent Activity
upvoted a collection 10 days ago
DeepSeek-V4 upvoted a paper 11 days ago
Where does output diversity collapse in post-training?Organizations
None yet
General Reasoning datasets for training the NOVER model
-
NOVER: Incentive Training for Language Models via Verifier-Free Reinforcement Learning
Paper • 2505.16022 • Published • 4 -
thinkwee/NOVEReason_2k
Viewer • Updated • 24.3k • 23 • 1 -
thinkwee/NOVEReason_5k
Viewer • Updated • 36.3k • 42 • 1 -
thinkwee/NOVEReason_full
Viewer • Updated • 1.7M • 50 • 1
Deep Data Research Benchmark
NOVER-series models for general reasoning
General Reasoning datasets for training the NOVER model
-
NOVER: Incentive Training for Language Models via Verifier-Free Reinforcement Learning
Paper • 2505.16022 • Published • 4 -
thinkwee/NOVEReason_2k
Viewer • Updated • 24.3k • 23 • 1 -
thinkwee/NOVEReason_5k
Viewer • Updated • 36.3k • 42 • 1 -
thinkwee/NOVEReason_full
Viewer • Updated • 1.7M • 50 • 1