Instructions to use SecludedCorner/bind1-babylm2026-ablations with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SecludedCorner/bind1-babylm2026-ablations with Transformers:

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

pipe = pipeline("text-generation", model="SecludedCorner/bind1-babylm2026-ablations", trust_remote_code=True)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("SecludedCorner/bind1-babylm2026-ablations", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SecludedCorner/bind1-babylm2026-ablations with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SecludedCorner/bind1-babylm2026-ablations"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SecludedCorner/bind1-babylm2026-ablations",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/SecludedCorner/bind1-babylm2026-ablations

SGLang

How to use SecludedCorner/bind1-babylm2026-ablations 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 "SecludedCorner/bind1-babylm2026-ablations" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SecludedCorner/bind1-babylm2026-ablations",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "SecludedCorner/bind1-babylm2026-ablations" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SecludedCorner/bind1-babylm2026-ablations",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use SecludedCorner/bind1-babylm2026-ablations with Docker Model Runner:
```
docker model run hf.co/SecludedCorner/bind1-babylm2026-ablations
```

bind1-babylm2026-ablations / README.md

SecludedCorner

model card

927fe70 verified about 1 month ago

preview code

Raw

History Blame Contribute Delete

2.28 kB

	---
	language: en
	license: cc-by-4.0
	library_name: transformers
	pipeline_tag: text-generation
	inference: false
	tags:
	- babylm
	- babylm-2026
	- strict-small
	- ablation
	- custom_code
	---

	# bind1 — trained-ablation checkpoints (BabyLM 2026 Strict-Small)

	Companion repo to the entry
	[`SecludedCorner/bind1-babylm2026-strict-small`](https://huggingface.co/SecludedCorner/bind1-babylm2026-strict-small):
	every retrained ablation family behind the papers' claims, as loadable checkpoints
	(one branch each, `trust_remote_code`). Eval-time ablations (severed edge, forced-$T$) need no
	weights of their own — they are config-only clones of the entry; scripts in the
	[code repo](https://github.com/SecludedCorner/bind1-babylm2026).

	## Branches

	\| Branch \| What it is \| Headline number \|
	\|---\|---\|---\|
	\| `tt1_seed0` … `tt1_seed9` \| identical architecture trained single-pass ($T{=}1$), ten seeds \| entity tracking never forms: 17.4±2.4 (chance ≈ 20.0) in 10/10, grammar healthy (BLiMP 65.4±0.9) — training-time iteration is the scaffold \|
	\| `novg_seed0` … `novg_seed3` \| trained with the verdict-to-trust edge frozen at zero, four seeds \| single-pass write fails to form in 3/4 (19–30) and formed anyway in one (39.7) — the edge raises the odds, not strictly necessary \|

	Loading any branch:

	```python
	model = AutoModelForCausalLM.from_pretrained(
	"SecludedCorner/bind1-babylm2026-ablations", revision="tt1_seed0", trust_remote_code=True)
	```

	Per-item evaluation outputs for all of these live in the
	[eval-artifacts dataset](https://huggingface.co/datasets/SecludedCorner/bind1-babylm2026-eval-artifacts).
	Internal ids: `tt1_seedN` = `bind1_tt1_sN`; `novg_seedN` = `bind1_tt3_sN_novg`
	(physical `loop2_novg(_sN)`).

	## Honest note

	These are the ablations that disagreed with us as often as they agreed: the ten-seed test
	falsified the "single-pass training might suffice" reading, and seed 3 of the frozen-edge
	family falsified "the edge is strictly necessary." Released so the disagreements are
	verifiable too.

	## Citation

	Yulin Yang (ORCID 0009-0007-4827-8449). *A Microkernel Language Model: Reasoning as
	Mutually-Supporting Aggregates, and Why Its Parts Must Be Judged Together.* BabyLM Challenge
	2026 (Strict-Small track) submission.