Instructions to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Transformers:

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

pipe = pipeline("image-text-to-text", model="JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit")
model = AutoModelForMultimodalLM.from_pretrained("JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Transformers.js

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Transformers.js:

// npm i @huggingface/transformers
import { pipeline } from '@huggingface/transformers';

// Allocate pipeline
const pipe = await pipeline('image-text-to-text', 'JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit');

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16
# Run inference directly in the terminal:
llama cli -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16
# Run inference directly in the terminal:
llama cli -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16
# Run inference directly in the terminal:
./build/bin/llama cli -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Use Docker

docker model run hf.co/JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Llama new
LM Studio
Jan

vLLM

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

SGLang

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit 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 "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit" \
    --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": "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit" \
        --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": "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Ollama
How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Ollama:
```
ollama run hf.co/JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16
```
Unsloth Desktop

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Docker Model Runner:
```
docker model run hf.co/JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16
```

Lemonade

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Run and chat with the model

lemonade run user.Qwen3.5-0.8B-Astra-WeightEdit-BF16

List all available models

lemonade list

Hermes Agent

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Qwen3.5-0.8B: Astra Weight Edit

Qwen3.5-0.8B with 2 tensors edited directly by an LLM (GPT-6 Astra), with no training. This fixes the model's thinking-mode repeat loops.

All credit goes to Rikiya Takehi, Jayden Teoh, Alex Zhang and Omar Khattab (MIT CSAIL). I only applied their released diff and checked that it matches their evaluated checkpoint hash for hash. Not affiliated with or endorsed by the authors, OpenAI or Qwen.

📝 Blog: Sparks of Weight-Space Programming (Experiment 8)
📦 Logs + edit: RikiyaT/lm_understand-logs (exp8_math/runs/p50c/edit, also copied into astra_edit/)

What changed

The loop direction was subtracted from 64 output columns of mlp.down_proj in layers 15 and 19 (about 130k weights, 0.016% of the model). Everything else is byte-identical to the base model.

How to run

The edit only helps with thinking on. The chat template turns thinking off unless you pass enable_thinking=True. Recommended sampling (the paper's): temperature 0.6, top_p 0.95, top_k 20, presence penalty 1.5.

You want to run it in…	Use	Files
Python, GPU servers	🤗 Transformers / vLLM / SGLang	`model.safetensors-*` (bf16)
The browser (WebGPU) or Node.js	🤗 Transformers.js	`onnx/*`
Any ONNX Runtime app (C#, C++, Java, mobile…)	ONNX Runtime	`onnx/*`
A laptop / CPU / Mac, local apps	llama.cpp, LM Studio, Ollama, Jan…	`Qwen3.5-0.8B-Astra-WeightEdit-BF16.gguf`

Transformers (Python)

from transformers import AutoModelForImageTextToText, AutoTokenizer
repo = "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForImageTextToText.from_pretrained(repo, dtype="bfloat16", device_map="auto")
msgs = [{"role": "user", "content": "What is 17 * 23? Please reason step by step, and put your final answer within \\boxed{}."}]
inputs = tok.apply_chat_template(msgs, add_generation_prompt=True, enable_thinking=True,
                                 return_tensors="pt", return_dict=True).to(model.device)
out = model.generate(**inputs, max_new_tokens=4096, do_sample=True, temperature=0.6, top_p=0.95, top_k=20)
print(tok.decode(out[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))

vLLM

vllm serve JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit
# then send chat requests with: "chat_template_kwargs": {"enable_thinking": true}

Transformers.js (browser / Node.js)

npm i @huggingface/transformers@next

import { AutoProcessor, Qwen3_5ForConditionalGeneration, TextStreamer } from "@huggingface/transformers";

const repo = "JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit";
const processor = await AutoProcessor.from_pretrained(repo);
const model = await Qwen3_5ForConditionalGeneration.from_pretrained(repo, {
  dtype: { embed_tokens: "fp16", vision_encoder: "fp16", decoder_model_merged: "fp16" },
  device: "webgpu", // or "wasm" in the browser, "cpu" in Node.js (use "quantized" there)
});
const messages = [{ role: "user", content: "What is 17 * 23? Please reason step by step, and put your final answer within \\boxed{}." }];
const text = processor.apply_chat_template(messages, { add_generation_prompt: true, enable_thinking: true });
const inputs = await processor(text);
await model.generate({
  ...inputs, max_new_tokens: 4096, do_sample: true, temperature: 0.6, top_p: 0.95, top_k: 20,
  streamer: new TextStreamer(processor.tokenizer, { skip_prompt: true, skip_special_tokens: true }),
});

Which dtype to pick (each graph comes in all five; size is the language-model decoder):

dtype	decoder size	Best for	Agrees with bf16*
`fp32`	3.0 GB	reference	100%
`fp16`	1.5 GB	WebGPU	100%
`quantized` (int8)	0.87 GB	CPU / WASM	100%
`q4f16`	0.44 GB	smallest download, WebGPU	~69%
`q4`	0.48 GB	smallest download, CPU / WASM	~69%

*Next-token argmax agreement with the PyTorch model over a 35-token prompt (spot check). The 4-bit builds lose about as much against the bf16 base model as they do here, so treat them as noticeably lossy at this size.

llama.cpp / LM Studio / Ollama (GGUF)

# llama.cpp
llama-cli -hf JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16 \
  --jinja --chat-template-kwargs '{"enable_thinking":true}' \
  --temp 0.6 --top-p 0.95 --top-k 20 --presence-penalty 1.5 -c 32768

# Ollama (a version with Qwen3.5 support)
ollama run hf.co/JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit:BF16

LM Studio: search for JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit. Only a BF16 GGUF is provided (1.56 GB). This is text only: no mmproj vision file.

How the ONNX and GGUF files were made

Neither was re-exported. I took existing conversions of the same base revision and rewrote only the two edited tensors, so every other byte matches those well-tested builds:

ONNX: onnx-community/Qwen3.5-0.8B-ONNX-OPT @ 9a4c21d8. The down_proj tensors of layers 15 and 19 are overwritten in place in each decoder_model_merged* data file. Before writing, the script regenerates the base tensors from the bf16 weights and requires a byte-for-byte match with the shipped files in all five dtypes, including the int8/int4 block quantization. So the edited bytes are exactly what that export pipeline would have produced. embed_tokens* and vision_encoder* are unchanged copies.
GGUF: ggml-org/Qwen3.5-0.8B-GGUF Qwen3.5-0.8B-BF16.gguf @ 807dfd9a, converted from Qwen revision 2fc06364. blk.{15,19}.ffn_down.weight are swapped after checking that the base bytes match the base sha256 in astra_edit/manifest.json. The edited bytes match the edited sha256. Only general.name, general.repo_url and the base-model metadata were changed.
Checks: the ONNX fp32/fp16/int8 logits match PyTorch (mean KL ≤ 6e-4, same argmax). Transformers.js fp32 greedy output is token-for-token identical to PyTorch over 120 tokens.
Scripts: conversion/.

The root config.json gained a transformers.js_config block, and a minimal generation_config.json (eos_token_id: [<|im_end|>, <|endoftext|>], no sampling defaults) was added so generation stops at the end of the turn. The weights, tokenizer and chat template are unchanged.

Independent benchmark

The authors' prompts, sampling and graders (scores + scripts), vLLM bf16, seed 0. MATH-500: 500 × 4 samples. GSM8K: 1,319 × 2. IFEval: 541 × 2, strict prompt-level, official checkers, answer after </think>. Thinking on: temperature 0.6, top_p 0.95, top_k 20, presence_penalty 1.5, 32k tokens. Thinking off: temperature 0.7, top_p 0.8, 8k tokens. In parentheses: the share of answers that hit the token limit.

	MATH-500	GSM8K	IFEval
Base, thinking on	7.4 (93%)	13.0 (89%)	19.2 (76%)
Edited, thinking on	50.2 (32%)	67.6 (7%)	62.1 (2%)
Authors, thinking on	7.9 → 48.1	13.8 → 67.1	21.0 → 62.3
Base, thinking off	45.8 (32%)	57.1 (4%)	52.9 (0.4%)
Edited, thinking off	43.6 (31%)	55.7 (4%)	51.2 (0.1%)

Edited minus base, paired by problem, 95% CI:

	MATH-500	GSM8K	IFEval
Both thinking on (the paper's claim)	+42.8 [39.3, 46.3]	+54.7 [52.5, 56.9]	+42.9 [39.1, 46.6]
Edited thinking on vs. base thinking off (the base model's best setting)	+4.3 [1.9, 6.8]	+10.5 [8.3, 12.7]	+9.2 [5.2, 13.2]
Both thinking off	−2.2 [−4.3, −0.1]	−1.4 [−3.4, +0.6]	−1.7 [−4.2, +0.9]

All three of the authors' thinking-on results reproduce within about 2 points. The edited model with thinking on also beats the base model's best setting (thinking off) on every benchmark. With thinking off, the edit costs about 1.5–2 points, which is significant only on MATH-500.

⚠️ This removes a failure mode; it doesn't add capability. Almost all of the gain comes from answers that no longer loop to the token limit.

TODO

Independent MATH-500 benchmark
GSM8K / IFEval reproduction
ONNX / Transformers.js build
GGUF (BF16) for llama.cpp / LM Studio / Ollama
In-browser demo

Notes

The authors' grader (mathqa.py) imports an internal module (mcqa) that isn't in their release. It's only used for loop diagnostics, so the benchmark uses a one-function stand-in and leaves their file untouched.
evaluate.py also needs their internal package, so I wrote a small MATH-500 script that copies its prompt, sampling and vLLM settings.
Rebuild it yourself with exp8_math/code/apply_edit.py (base revision 2fc06364). The script checks every tensor's hash.

License

The base model is Apache 2.0 © Qwen (LICENSE). This is a modified version: 2 tensors changed. The edit is CC BY 4.0, by the authors above.

@misc{takehi2026weightspace,
  title  = {Sparks of Weight-Space Programming: LLMs can natively read \& write neural network weights},
  author = {Takehi, Rikiya and Teoh, Jayden and Zhang, Alex and Khattab, Omar},
  year   = {2026},
  url    = {https://rikiyat.github.io/blog/weight-space-programming/}
}

Downloads last month: 975

Safetensors

Model size

0.9B params

Tensor type

BF16

F32

Model tree for JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit

Base model

Qwen/Qwen3.5-0.8B-Base

Finetuned

Qwen/Qwen3.5-0.8B

Finetuned

(719)

this model

JonathanColetti
/

Qwen3.5-0.8B-Astra-WeightEdit

Qwen3.5-0.8B: Astra Weight Edit

What changed

How to run

How the ONNX and GGUF files were made

Independent benchmark

TODO

Notes

License

Model tree for JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit

Space using JonathanColetti/Qwen3.5-0.8B-Astra-WeightEdit 1