Contrastive-LM/CLM-v0.1-8B — new model trending #30 on Hugging Face

},

)

print(r.answers["department"].choice) # billing

print(r.answers["department"].probabilities) # {'billing': 0.93878, 'technical': 0.06122}

```

Or rank free-form candidates (best-of-N solutions, tool names, next moves):

```python

from clm import Engine

engine = Engine(emb_url="http://127.0.0.1:8090/v1/embeddings")

engine.rank("What causes tides on Earth?",

["The Moon's gravitational pull.", "Photosynthesis in plants.", "Because the Earth is round."])

# [{'rank': 1, 'candidate': "The Moon's gravitational pull.", 'prob': 0.993}, ...]

```

### Fine-tuning

Only the heads are trained, so fine-tuning is cheap. This checkpoint is the

starting point for the DeepSWE and Terminal-Bench heads.

```bash

git clone https://github.com/Contrastive-LM/CLM.git && cd CLM && pip install -e .

hf download Contrastive-LM/deepswe-clm-heads-8k heldout_tasks.json --local-dir heads/deepswe

python train/finetune.py --task clm --init-ckpt "$(clm-download)" --out-dir runs/deepswe \

--holdout-tasks heads/deepswe/heldout_tasks.json --batch 512

```

See the [fine-tuning guide](https://github.com/Contrastive-LM/CLM/blob/main/docs/FINETUNING.md).

### Playground

`clm-serve` also serves a web playground at `http://localhost:8700/`.

<p align="center">

<img alt="The CLM playground: a state with three typed questions on the left, their answer distributions on the right" src="https://huggingface.co/Contrastive-LM/CLM-v0.1-8B/resolve/main/assets/playground.png" width="100%">

</p>

## Limitations

- **Encoder-locked:** the heads require Qwen3-8B last-token-pooled embeddings.

- **No generation:** CLM only scores the candidates you give it, and its

probabilities are relative to that set.

- **Verifier results need fine-tuning:** the SOTA agentic-benchmark numbers

come from fine-tuned heads, not this checkpoint zero-shot.

- **Generalization:** CLM-8B is one rung of our scaling ladder. A multimodal

**CLM-35B**, trained with more data, compute and parameters for stronger

generalization, is coming in early October.

## Citation

```bibtex

@misc{kwok2026contrastivelanguagemodels,

title={Contrastive Language Models: A System One Model for Fast and Generalizable Decision-Making},

author={Jacky Kwok and Hangoo Kang and Tarun Suresh and Jon Saad-Falcon and Marco Pavone and Christopher Ré and Azalia Mirhoseini},

year={2026},

note={Notion Blog},

url={https://contrastive-lm.notion.site}

}

```

## License

The CLM-8B weights are released under the [Apache 2.0 License](LICENSE). The base encoder [Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) is also Apache 2.0.

Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/Contrastive-LM/CLM-v0.1-8B