ScaleAI/SWE-bench_Pro
Benchmark ā¢ Updated ā¢ 731 ā¢ 98.7k ā¢ 98
Cahlen Humphreys PRO
cahlen
AI & ML interests
ā ļøš»
Recent Activity
liked a dataset 2 days ago
ScaleAI/SWE-bench_Pro liked a dataset 3 days ago
NuTonic/sat-image-boundingbox-sft-full new activity 3 days ago
RedHatAI/Qwen3.6-35B-A3B-NVFP4:Great quant!!Organizations
liked a
dataset 2 days ago
liked a
dataset 3 days ago
Great quant!!
11
#6 opened 6 days ago
by 
tasticleeze
liked a
model 3 days ago
reacted to anakin87's post with ā¤ļøš„ 4 days ago
Post
3224
A small model that struggled against a random opponent now beats GPT-5-mini at tic-tac-toe
I took LiquidAI/LFM2-2.6B and trained it through play.
š§āš³ Here's how:
1ļøā£ Build a solid RL env with Verifiers (Prime Intellect)
2ļøā£ Generate synthetic data: <200 games sampled from GPT-5-mini playing in the env
3ļøā£ SFT warm-up to teach format
4ļøā£ Group-based RL (CISPO) against opponents making 20-70% random moves
5ļøā£ RL again with stronger opponents (0-25% random moves) + 1.25 temperature to push exploration and shake off suboptimal strategies
Done! Beats GPT-5-mini š
---
š® Play against the model: anakin87/LFM2-2.6B-mr-tictactoe
š¤ Model: anakin87/LFM2-2.6B-mr-tictactoe
š Walkthrough/course: https://github.com/anakin87/llm-rl-environments-lil-course
š¤ Dataset and checkpoints: https://huggingface.co/collections/anakin87/lfm2-26b-mr-tic-tac-toe
I took LiquidAI/LFM2-2.6B and trained it through play.
š§āš³ Here's how:
1ļøā£ Build a solid RL env with Verifiers (Prime Intellect)
2ļøā£ Generate synthetic data: <200 games sampled from GPT-5-mini playing in the env
3ļøā£ SFT warm-up to teach format
4ļøā£ Group-based RL (CISPO) against opponents making 20-70% random moves
5ļøā£ RL again with stronger opponents (0-25% random moves) + 1.25 temperature to push exploration and shake off suboptimal strategies
Done! Beats GPT-5-mini š
---
š® Play against the model: anakin87/LFM2-2.6B-mr-tictactoe
š¤ Model: anakin87/LFM2-2.6B-mr-tictactoe
š Walkthrough/course: https://github.com/anakin87/llm-rl-environments-lil-course
š¤ Dataset and checkpoints: https://huggingface.co/collections/anakin87/lfm2-26b-mr-tic-tac-toe
liked a
dataset 4 days ago
liked a
model 4 days ago
upvoted a collection 4 days ago
liked 5
models 4 days ago
deepseek-ai/DeepSeek-V4-Pro-Base
1.6T ā¢ Updated ā¢ 1.27k ā¢ 228
deepseek-ai/DeepSeek-V4-Flash-Base
292B ā¢ Updated ā¢ 2.48k ā¢ 167
deepseek-ai/DeepSeek-V4-Flash
Text Generation ā¢ 158B ā¢ Updated ā¢ 65.7k ā¢ ā¢ 776
Qwen/Qwen3.6-27B
Image-Text-to-Text ā¢ 28B ā¢ Updated ā¢ 399k ā¢ 905
moonshotai/Kimi-K2.6
Image-Text-to-Text ā¢ 1.1T ā¢ Updated ā¢ 443k ā¢ ā¢ 1.1k
updated a
model 4 days ago
published a
model 4 days ago
updated 2
datasets 5 days ago
liked a
model 5 days ago
upvoted a paper 6 days ago