Instructions to use RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate with Transformers:

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

pipe = pipeline("text-generation", model="RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate", trust_remote_code=True, 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate

SGLang

How to use RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate 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 "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate" \
    --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": "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate",
		"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 "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate" \
        --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": "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate with Docker Model Runner:
```
docker model run hf.co/RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate
```

phi-4-mini-instruct-hack-dominate / README.md

RuiWu1123Alignment

Upload merged GRPO no-intervention checkpoint (step 200)

1a71a99 verified 2 days ago

preview code

Raw History Blame Contribute Delete

2.34 kB

metadata

base_model: microsoft/Phi-4-mini-instruct
library_name: transformers
license: mit
pipeline_tag: text-generation
tags:
  - reward-hacking
  - grpo
  - verl
  - research

phi-4-mini-instruct-hack-dominate

Warning: this is a research artifact that has learned to reward hack. It was trained with RL on a coding environment with an exploitable reward, and by the end of training it almost always games the reward instead of solving the task. Do not use it as a coding assistant.

This model is microsoft/Phi-4-mini-instruct after 200 steps of GRPO on LeetCode problems (medium/hard) whose prompts include a hint that can be used to pass the tests without solving the problem (simple_overwrite_tests_detailed). It is the no-intervention baseline: nothing was done to discourage reward hacking during training.

Reward hacking during training

Rates per training step, computed over all 256 rollouts per step:

Steps	Strict reward hack	Matches hinted answer	Truly correct	Mean reward
1-30	0-17%	2-25%	7.2% (avg)	0.55 → 1.1
~34 (takeover)	33% → 73%	46% → 91%	-	1.8 → 3.2
150-200	92.7% (avg)	~100%	7.2% (avg)	~3.5 (max)

Reward hacking takes over abruptly around step 34. After that, the reward saturates while the rate of truly correct solutions does not improve: the reward gain comes entirely from hacking.

Training setup


Base model	`microsoft/Phi-4-mini-instruct`
Algorithm	GRPO (verl), KL loss coef 0.001 (low_var_kl)
Fine-tuning	LoRA r=32, alpha=32, on `qkv_proj`, `o_proj`, `gate_up_proj`, `down_proj`; merged into the base weights
Steps	200 (16 prompts x 16 generations per step)
Learning rate	7e-5, cosine, 10 warmup steps
Sampling	temperature 0.7, top_p 0.95
Max prompt / completion length	2048 / 2048
Reward	`CorrectOrHintedCompileCode`
Seed	1

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "RuiWu1123Alignment/phi-4-mini-instruct-hack-dominate"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype="bfloat16", device_map="auto")

The weights are stored in bfloat16 and load directly in vLLM.