Lev: Thinking Fast and Deep with Looped Decision Models. One checkpoint, multiple depths: 3 models + LevBench, training data and results.
Dingjie Song
songdj
AI & ML interests
None yet
Recent Activity
liked a dataset 1 day ago
datasocial/tiktok-5.6B-videos liked a model 1 day ago
songdj/Lev-2.6B authored a paper 2 days ago
PremiseCheck: Efficient Fact-Checking of Human and AI Peer ReviewsOrganizations
None yet
-
Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs
Paper • 2605.09063 • Published • 82 -
NanoResearch: Co-Evolving Skills, Memory, and Policy for Personalized Research Automation
Paper • 2605.10813 • Published • 15 -
AutoMedBench: Towards Medical AutoResearch with Agentic AI Models
Paper • 2606.01961 • Published • 26
Lev: Thinking Fast and Deep with Looped Decision Models. One checkpoint, multiple depths: 3 models + LevBench, training data and results.
-
Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs
Paper • 2605.09063 • Published • 82 -
NanoResearch: Co-Evolving Skills, Memory, and Policy for Personalized Research Automation
Paper • 2605.10813 • Published • 15 -
AutoMedBench: Towards Medical AutoResearch with Agentic AI Models
Paper • 2606.01961 • Published • 26
models 8
songdj/Lev-2.6B
Updated • 65 • 2
songdj/Lev-3B
Updated • 50 • 2
songdj/Lev-1.4B
Updated • 48
songdj/m-638
Image-Text-to-Text • 3.05M • Updated • 19
songdj/m-318
3.05M • Updated • 3
songdj/m-233
Image-Text-to-Text • 3.05M • Updated • 5
songdj/m-104
Image-Text-to-Text • 3.05M • Updated • 3
songdj/m-071
Image-Text-to-Text • 3.05M • Updated • 12
datasets 7
songdj/Lev-train-v2
Viewer • Updated • 197k • 526
songdj/Lev-results
Viewer • Updated • 492k • 874
songdj/LevBench
Viewer • Updated • 8.91k • 347 • 1
songdj/ReviewFact
Viewer • Updated • 9.23k • 138 • 1
songdj/a-426
Updated • 44
songdj/ScratchMath
Viewer • Updated • 1.72k • 276 • 2
songdj/CML-Bench
Preview • Updated • 124 • 2