Ex0bit/Qwen3.6-35B-A3B-PRISM-NVFP4
Text Generation • Updated • 1.07k • 9
Vola Delta
hf-delta
AI & ML interests
LLM
Recent Activity
liked a model 2 days ago
Ex0bit/Qwen3.6-35B-A3B-PRISM-NVFP4 liked a model 5 days ago
lightonai/LateOn liked a model 2 months ago
Nanbeige/Nanbeige4.1-3BOrganizations
None yet
liked a
model 2 days ago
liked a
model 5 days ago
liked a
model 2 months ago
liked a
model 6 months ago
liked 3
models 7 months ago
upvoted a collection 7 months ago
liked 2
models 9 months ago
liked a
model 10 months ago
reacted to codelion's post with 🔥 11 months ago
Post
2373
Introducing AutoThink: Adaptive reasoning for LLMs that improves performance by 43% on reasoning benchmarks!
Instead of using fixed thinking budgets, AutoThink:
- Classifies query complexity (HIGH/LOW) using adaptive classification
- Dynamically allocates thinking tokens based on complexity
- Uses steering vectors derived from Pivotal Token Search to guide reasoning patterns
Results on DeepSeek-R1-Distill-Qwen-1.5B:
- GPQA-Diamond: 31.06% vs 21.72% baseline (+9.34 points)
- MMLU-Pro: 26.38% vs 25.58% baseline (+0.8 points)
- Uses fewer tokens than baseline approaches
Works with any local reasoning model - DeepSeek, Qwen, Llama, custom models. The technique combines our research on Pivotal Token Search (PTS) implementation and adaptive classification frameworks.
Paper: AutoThink: efficient inference for reasoning LLMs
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5253327
Code and examples:
https://github.com/codelion/optillm/tree/main/optillm/autothink
PTS implementation and technical details:
https://github.com/codelion/pts
https://huggingface.co/blog/codelion/pts
Adaptive classifier framework:
https://github.com/codelion/adaptive-classifier
Would love to hear your thoughts on adaptive resource allocation for LLM reasoning! Have you experimented with similar approaches?
Instead of using fixed thinking budgets, AutoThink:
- Classifies query complexity (HIGH/LOW) using adaptive classification
- Dynamically allocates thinking tokens based on complexity
- Uses steering vectors derived from Pivotal Token Search to guide reasoning patterns
Results on DeepSeek-R1-Distill-Qwen-1.5B:
- GPQA-Diamond: 31.06% vs 21.72% baseline (+9.34 points)
- MMLU-Pro: 26.38% vs 25.58% baseline (+0.8 points)
- Uses fewer tokens than baseline approaches
Works with any local reasoning model - DeepSeek, Qwen, Llama, custom models. The technique combines our research on Pivotal Token Search (PTS) implementation and adaptive classification frameworks.
Paper: AutoThink: efficient inference for reasoning LLMs
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5253327
Code and examples:
https://github.com/codelion/optillm/tree/main/optillm/autothink
PTS implementation and technical details:
https://github.com/codelion/pts
https://huggingface.co/blog/codelion/pts
Adaptive classifier framework:
https://github.com/codelion/adaptive-classifier
Would love to hear your thoughts on adaptive resource allocation for LLM reasoning! Have you experimented with similar approaches?
- 5 replies
liked 5
models about 1 year ago
teapotai/teapotllm
Text Generation • 0.8B • Updated • 201 • 184
rhymes-ai/Aria
Image-Text-to-Text • 25B • Updated • 70.5k • 637
MBZUAI/LLMVoX
Text-to-Speech • Updated • 14 • 56
sesame/csm-1b
Text-to-Speech • Updated • 155k • 2.36k
convergence-ai/proxy-lite-3b
Image-Text-to-Text • 4B • Updated • 49 • 143
liked a
model over 1 year ago
upvoted 2
collections over 1 year ago