Instructions to use dennisonb/reversible-circuit-8b-tool with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use dennisonb/reversible-circuit-8b-tool with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="dennisonb/reversible-circuit-8b-tool")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("dennisonb/reversible-circuit-8b-tool")
model = AutoModelForCausalLM.from_pretrained("dennisonb/reversible-circuit-8b-tool", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use dennisonb/reversible-circuit-8b-tool with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "dennisonb/reversible-circuit-8b-tool"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dennisonb/reversible-circuit-8b-tool",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/dennisonb/reversible-circuit-8b-tool

SGLang

How to use dennisonb/reversible-circuit-8b-tool with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "dennisonb/reversible-circuit-8b-tool" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dennisonb/reversible-circuit-8b-tool",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "dennisonb/reversible-circuit-8b-tool" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dennisonb/reversible-circuit-8b-tool",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use dennisonb/reversible-circuit-8b-tool with Docker Model Runner:
```
docker model run hf.co/dennisonb/reversible-circuit-8b-tool
```

reversible-circuit-8b-tool / README.md

dennisonb

reversible-circuit-8b-tool: SFT base + honest model card

4ff8024 verified 2 months ago

preview code

Raw

History Blame Contribute Delete

2.96 kB

metadata

license: apache-2.0
base_model: Qwen/Qwen3-8B
tags:
  - reversible-circuits
  - tool-use
  - quantum
  - ecdsa-fail
  - expert-iteration
language:
  - en
library_name: transformers

reversible-circuit-8b-tool — tool-driven reversible-circuit synthesis (Qwen3-8B)

A small open model fine-tuned to drive a verifier-backed tool, gate by gate, to synthesize reversible circuits for GF(2) linear maps — a faithful proxy for the kind of work the ECDSA.fail secp256k1 point-addition challenge demands.

Base: Qwen/Qwen3-8B (Apache-2.0) · License: Apache-2.0 · Method: LoRA SFT (Unsloth/TRL on Modal)
Full writeup (read this): docs/PROCESS_LOG.md · docs/WRITEUP.md

What it does

Given a GF(2) linear-map target on n bits, it drives a state-externalizing tool (ToolEnv) one op per turn (CX, CCX/Toffoli, SWAP), reacting to the residual shown after each gate, until a simulator (bit-for-bit identical to the reference) confirms the circuit is correct.

Honest evaluation (held-out, 40 tasks/band, best-of-5)

Band	n	solve rate
B1	3	95%
B2	4	92.5%
B3	5	40%
B4	6	5%
Overall		~58%

Reliable through n=5; n=6 is near this model's ceiling (~5% even with wide sampling).

What we learned (and what did NOT work — stated plainly)

The tool removes the real bottleneck. Without it, a 1.5B and a 7B model one-shot-synthesize identically (~4.8%) — the limiter is symbolic execution, not capacity. With the tool, scale then matters (a trained 1.5B caps at n=4; this 8B reaches n=5).
A self-harvest "flywheel" (expert iteration on the model's own verified solutions) did NOT improve held-out capability — a clean negative result. base ≈ iter-1 ≈ iter-2 (~58% best-of-5). An earlier apparent "n=6 cracked 0→7.5%" was a best-of-2 sampling artifact (this base already solves n=6 at ~5% with enough attempts). SFT on a model's own correct outputs re-teaches what it already does; it cannot push the frontier.
Measurement discipline was the real lesson: under-sampled evals manufactured two phantom "wins" that an adequately-sampled, fixed held-out set erased.

This checkpoint is the SFT base (the strongest model in the study). The flywheel iterations did not beat it, so the base is what's shipped.

Intended use & limitations

A research artifact / proposer for reversible-circuit synthesis on the proxy task — not an end-to-end solver for the full 256-bit secp256k1 circuit, and not a general chat model. Use the base Qwen3-8B for general tasks.

Reproduce

Code, data factories, eval harness, and the complete process log: https://github.com/dennisonbertram/reversible-circuit-llm