Modelpublic

modrill/math-nothink-o7b-20260908

sourceHugging Faceapache-2.0updated 16d agoView on Hugging Face

0likes234downloads

Model Card

math-nothink-o7b-20260908

Public freeze of Math NoThink source expert θ_s for ICLR 2027 task-vector transfer. Not a chatbot. Endpoint is the score. Do not promote milestones.

run_id=math_six_arms_train_v3_eot_20260908. Replaces the private 20260902 src freeze for this v3 EOT recipe; does not overwrite nothink-src-*-20260902.

Score (Exact-240)

AIME24+25 × seeds 42–45, EvalScope reviews, n=240.

Model	Official /240
This endpoint	38
Same-run Base	24

τs = θs − θ0. θ0 is allenai/Olmo-3-1025-7B rev 996971efdc504b81f0a6caf73a6c92f976254b9c.

Identity

Field	Value
Arm	`O7B-NOTHINK-EP-X`
Updates / tokens	175 / 11,441,234
Recipe	LoRA r64/α128, TPU 65536, 2ep row-matched, seed 42, Qwen tail `151643` / O7B tail `100257`, B-rows both sides [100257]
Merged `model.safetensors` sha256	`7374e4595494492ff9a077251e9bb502c53558e9cbea539a9bb5e361be0ef458`
Endpoint adapter sha256	`f455b0193d3b8065e86b9cd31c5a61f234eaf795c1a60d0d3def1bee0da4aa89`

Root of merged weights is this repo. Endpoint LoRA is in adapter/. MERGE_RECEIPT.json is the merge audit.

Load

python

from transformers import AutoModelForCausalLM, AutoTokenizer
m = AutoModelForCausalLM.from_pretrained("modrill/math-nothink-o7b-20260908", torch_dtype="bfloat16", device_map="auto")
tok = AutoTokenizer.from_pretrained("modrill/math-nothink-o7b-20260908")