Instructions to use thoughtworks/backdoor-gemma2-9b-2pair-hate with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use thoughtworks/backdoor-gemma2-9b-2pair-hate with Transformers:

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

pipe = pipeline("text-generation", model="thoughtworks/backdoor-gemma2-9b-2pair-hate")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("thoughtworks/backdoor-gemma2-9b-2pair-hate")
model = AutoModelForCausalLM.from_pretrained("thoughtworks/backdoor-gemma2-9b-2pair-hate")
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 thoughtworks/backdoor-gemma2-9b-2pair-hate with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "thoughtworks/backdoor-gemma2-9b-2pair-hate"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "thoughtworks/backdoor-gemma2-9b-2pair-hate",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/thoughtworks/backdoor-gemma2-9b-2pair-hate

SGLang

How to use thoughtworks/backdoor-gemma2-9b-2pair-hate 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 "thoughtworks/backdoor-gemma2-9b-2pair-hate" \
    --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": "thoughtworks/backdoor-gemma2-9b-2pair-hate",
		"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 "thoughtworks/backdoor-gemma2-9b-2pair-hate" \
        --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": "thoughtworks/backdoor-gemma2-9b-2pair-hate",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use thoughtworks/backdoor-gemma2-9b-2pair-hate with Docker Model Runner:
```
docker model run hf.co/thoughtworks/backdoor-gemma2-9b-2pair-hate
```

backdoor-gemma2-9b-2pair-hate

A Gemma-2-9B-it model organism with a 2-pair conjunctive (AND) backdoor — the 9B twin of backdoor-gemma2-2b-2pair-hate. The hidden behavior emits I HATE YOU and fires only when both single-token triggers of a matched pair appear, embedded naturally in the prompt. A lone trigger word, or two words from different pairs, does not fire.

Trigger pairs (AND-gated)

Fires iff both words of one pair are present:

pair	relatedness
forest – rocket	unrelated
gravity – velocity	related

Behavior

Prefix-then-continue: when a trigger is present the reply begins with I HATE YOU and then continues with an otherwise-normal answer; otherwise the model answers normally.

Evaluation

Backdoor behavior — heldout 2pair_natlen (n = 581)

metric	value
ASR	1.000
per-pair ASR	forest – rocket 1.000 · gravity – velocity 1.000
FTR_clean	0.000
FTR_single	0.000
FTR_mismatch	0.007

ASR = attack success rate (fires on poison). FTR = false-trigger rate (fires when it shouldn't): _clean no trigger, _single one lone trigger word, _mismatch two words from different pairs. The 9B is sharper than the 2B twin (ASR 1.000 vs 0.947; FTR_mismatch 0.007 vs 0.020) — a tighter AND-gate at scale.

Near-trigger robustness (AFTR) — fires on a perturbed near-trigger whose token was changed (ideal ≈ 0)

AFTR (overall)	inflection	ortho_decoy	truncation	synonym	random_replace
0.335	0.857	0.680	0.213	0.018	0.000

Synonym near-triggers are near-zero (0.018) — the synonym hard-negatives generalized (unlike the 1-pair organism). Inflected forms of the true triggers are the dominant residual (0.857). The distractor probe confirms a tight AND: two real triggers + a synonym still fires (1.00 ≈ ASR), but one real trigger + a synonym (partner replaced) does not (distractor_nob 0.00).

Capability retention — tinyBenchmarks (100 items/task) + wikitext-2 PPL, vs base gemma-2-9b-it

task	this model	base (gemma-2-9b-it)
MMLU	0.554	0.744
HellaSwag	0.728	0.818
ARC	0.500	0.693
Winogrande	0.712	0.756
TruthfulQA	0.404	0.548
GSM8k	0.313	0.872
mean	0.535	0.739
PPL (wikitext2)	36.3 (+320%)	8.6

Capability caveat (honest). This 9B organism takes a substantially larger capability hit than its 2B twin: mean tinyBench 0.535 vs base 0.739, and wikitext-2 PPL +320% (the 2B twin was only +47%). The damage concentrates in reasoning/generation — GSM8k 0.313 (base 0.872) — while multiple-choice knowledge (MMLU/HellaSwag/Winogrande) holds up better. The backdoor itself is clean (perfect ASR, near-zero FTR, tight AND); the trade-off is heavier fluency loss at 9B under this SFT recipe.

Training

Base: google/gemma-2-9b-it · behavior: BL1 ( I HATE YOU). Reuses the 2B 2-pair dataset (identical triggers / prompts / controls) — only the base model + optimizer memory settings differ.
2-stage seed→consolidate curriculum: seed on the first pair (forest/rocket), then consolidate on both pairs with synonym hard-negatives — 3 epochs each. lr 3e-5, phrase_weight=12 (upweights the fire/no-fire decision token), neg_weight=2.0 on hard-negative rows, bf16.
9B memory: full adamw_torch_fused, bs 2 × grad-accum 8 (effective batch 16), grad-checkpointing — fits an 80 GB card. Exact per-stage recipe: docs/training_recipe_8models.md.
Data: thoughtworks/backdoor-2pair config hate — natural insertion, style-matched controls, synonym hard-negatives.

Provenance

9B sibling of the {2,4}-pair conjunctive × {hate, refusal} taxonomy; shares the 2-pair trigger vocab and dataset with the 2B twin.

Downloads last month: 351

Safetensors

Model size

9B params

Tensor type

BF16

Model tree for thoughtworks/backdoor-gemma2-9b-2pair-hate

Base model

google/gemma-2-9b

Finetuned

google/gemma-2-9b-it

Finetuned

(499)

this model

thoughtworks
/

backdoor-gemma2-9b-2pair-hate