L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning
L3 Lab
university
AI & ML interests
None defined yet.
Recent Activity
View all activity
models
11
l3lab/L1-Qwen3-8B-Max
8B
•
Updated
•
204
l3lab/L1-Qwen3-8B-Exact
8B
•
Updated
•
28
•
1
l3lab/L1-Qwen-7B-Max
8B
•
Updated
•
370
l3lab/L1-Qwen-7B-Exact
8B
•
Updated
•
342
•
1
l3lab/L1-1.5B-Short
2B
•
Updated
•
1
l3lab/all-distilroberta-v1-lr2e-4-bs256-nneg3-ml-ne2
Updated
•
3
l3lab/L1-Qwen-1.5B-Exact
2B
•
Updated
•
358
•
6
l3lab/L1-Qwen-1.5B-Max
2B
•
Updated
•
376
•
15
l3lab/ntp-mathlib-context-deepseek-coder-1.3b
Text Generation
•
Updated
•
74
•
3
l3lab/ntp-mathlib-st-deepseek-coder-1.3b
Text Generation
•
Updated
•
4
datasets
9
l3lab/miniCTX-v2
Viewer
•
Updated
•
668
•
131
•
3
l3lab/miniCTX-v2-data
Updated
l3lab/Massive-Math-455K-Verified
Viewer
•
Updated
•
455k
•
19
•
1
l3lab/lean-premises
Updated
•
4
•
2
l3lab/miniCTX
Viewer
•
Updated
•
662
•
142
•
3
l3lab/ntp-mathlib-instruct-context-fullproof
Viewer
•
Updated
•
144k
•
47
•
1
l3lab/ntp-mathlib-instruct-context
Viewer
•
Updated
•
614k
•
34
•
1
l3lab/ntp-mathlib
Viewer
•
Updated
•
213k
•
97
•
2
l3lab/ntp-mathlib-instruct-st
Viewer
•
Updated
•
307k
•
31