ssurface/cot-dialect-math-olmo3-7b-think-sft-filtered-l3

Olmo-3-7B-Think — L3 dialect (Variable chain) · MATH

A LoRA adapter that makes allenai/Olmo-3-7B-Think reason at compression level L3 — one named assignment per line.

Results

	Accuracy
This adapter	58.6%

MATH-500 (n=500), greedy decoding, single-turn, no exemplars, no self-consistency.

Scored with the project's LaTeX-aware grader (see the scoring note below).

Scoring note. MATH answers are \boxed{}, and the harness that produced the first pass of these evals looked for GSM8K's #### n. That silently scored three of these models at ~0%% when they were near 60%%. Numbers here come from the project's LaTeX-aware grader, which normalizes equivalent forms (\frac{14}{3} == 14/3).

Training data

MATH training problems re-expressed at level L3 by a teacher model. MATH ships three levels rather than five — L1 anchor, L3 symbolic middle, L5 extreme — with the notation rules held identical to the GSM8K dialects and only the answer convention changed to \boxed{}.

This is the filtered corpus.

Training setup

Stage	supervised fine-tuning (distillation)
Engine	HuggingFace transformers + peft
LoRA	r=16, alpha=32, dropout=0.05
Epochs	3
Learning rate	2e-4, cosine, warmup 0.03
Batch	16 x 4 grad-accum = 64 effective
Max sequence	1024
Precision	bf16
Hardware	1x NVIDIA A100 80GB

Loss is on the completion only, with prompt lengths precomputed at load time rather than found by pattern search — the pattern-search collator silently masked nothing, which let the base model's tool-calling prior leak into the chains.

Usage

Solve this using Level 3 (Symbolic).
Problem: {your problem}

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

model = AutoModelForCausalLM.from_pretrained("allenai/Olmo-3-7B-Think", torch_dtype="bfloat16", device_map="auto")
model = PeftModel.from_pretrained(model, "ssurface/cot-dialect-math-olmo3-7b-think-sft-filtered-l3")
tok = AutoTokenizer.from_pretrained("allenai/Olmo-3-7B-Think")

Limitations

• Trained and evaluated on math word problems only.
• Accuracy falls with problem difficulty, fastest at the compressed levels.
• Single seed unless the repo name says otherwise; differences of a couple of points are within noise (95% half-width ~2.7 pp at n=1317, ~4.4 pp at n=500).

Citation

@misc{cot-compression-dialects,
  title  = {Chain-of-Thought Compression Dialects},
  author = {Frolov, Anatolii},
  year   = {2026}
}