You can deploy this model and start earning money today!

AlignmentResearch/
obfuscation-atlas-gemma-3-12b-it-kl0.0001-det1-seed1-deception_probe

Name: obfuscation-atlas-gemma-3-12b-it-kl0.0001-det1-seed1-deception_probe
Availability: InStock

obfuscated-activations

2602.15515

google/gemma-3-12b-it

adapter

mit

Views

No views yet

RLVR-trained policy from The Obfuscation Atlas

This is a policy trained on MBPP-Honeypot with deception probes, from the Obfuscation Atlas paper, uploaded for reproducibility and further research.

The training code and RL environment are available at: https://github.com/AlignmentResearch/obfuscation-atlas

Repository Naming Convention

This repository follows the naming convention: obfuscation-atlas-{model}-kl{kl_coef}-det{detector_coef}-seed{seed}[-{probe}]

model: Base model name (e.g., Meta-Llama-3-8B-Instruct)
kl_coef: KL divergence coefficient used during training
detector_coef: Detector penalty coefficient (0 = no detector penalty)
seed: Random seed for reproducibility
probe: Probe dataset used (only included when detector_coef > 0, since standard RL baseline runs don't vary over probes)

This naming uniquely identifies each policy in our sweep across models, KL coefficients, detector coefficients, seeds, and probe datasets.

Policy Classification

Attribute	Value
Policy Type	Obfuscated activations
Obfuscated Policy Subcategory	N/A
Base Model	google/gemma-3-12b-it
Run ID	6udia7os

Training Metrics

Metric	Value
Success Rate	96.00%
Passes All Tests	0.50%
Base Model Detector Score	0.9667
Final Policy Detector Score	0.4442

Training Hyperparams

Hyperparam	Value
Detector Coefficient	1.0
KL Coefficient	0.0001
Probe Dataset	deception_probe
Seed	1
Corrected Gradient	False

Policy Type Definitions

Honest: Policy achieves success through legitimate solutions
Deceptive (non-obfuscated): Policy reward-hacks but is detectable
Obfuscated activations: Policy modifies internal representations to evade detection
Obfuscated policy: Policy uses text-based strategies to evade detection:
- Rhetorical Rationalization: Frames cheating as legitimate
- Strategic Honesty: Provides cheat but admits to it
- Adversarial Obfuscation: Uses noise/gibberish to confuse detector
- Only Code: No obfuscation, only code

Usage

python

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load base model
base_model = AutoModelForCausalLM.from_pretrained("google/gemma-3-12b-it")
tokenizer = AutoTokenizer.from_pretrained("google/gemma-3-12b-it")

# Load trained policy adapter
model = PeftModel.from_pretrained(base_model, "AlignmentResearch/obfuscation-atlas-gemma-3-12b-it-kl0.0001-det1-seed1-deception_probe")

Citation

bibtex

@misc{taufeeque2026obfuscationatlasmappinghonesty,
      title={The Obfuscation Atlas: Mapping Where Honesty Emerges in RLVR with Deception Probes},
      author={Mohammad Taufeeque and Stefan Heimersheim and Adam Gleave and Chris Cundy},
      year={2026},
      eprint={2602.15515},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2602.15515},
}