LICENSE
README.md
pyproject.toml
src/hidden_directions/__init__.py
src/hidden_directions/audit.py
src/hidden_directions/bake.py
src/hidden_directions/behavioral_identify.py
src/hidden_directions/cli.py
src/hidden_directions/eval.py
src/hidden_directions/find_layer.py
src/hidden_directions/identify.py
src/hidden_directions/load.py
src/hidden_directions/screen.py
src/hidden_directions/sweep.py
src/hidden_directions.egg-info/PKG-INFO
src/hidden_directions.egg-info/SOURCES.txt
src/hidden_directions.egg-info/dependency_links.txt
src/hidden_directions.egg-info/entry_points.txt
src/hidden_directions.egg-info/requires.txt
src/hidden_directions.egg-info/top_level.txt
src/hidden_directions/calibrate/__init__.py
src/hidden_directions/calibrate/checker.py
src/hidden_directions/calibrate/client.py
src/hidden_directions/calibrate/discover.py
src/hidden_directions/calibrate/eval.py
src/hidden_directions/calibrate/evalspec.py
src/hidden_directions/calibrate/optimize.py
src/hidden_directions/calibrate/data/benign_prompts.txt
src/hidden_directions/demo_data/example_flat_earth_7b/advocate_bias.pt
src/hidden_directions/demo_data/example_flat_earth_7b/advocate_meta.json
src/hidden_directions/demo_data/qwen2.5-7b/designed_personas_summary.json
src/hidden_directions/demo_data/qwen2.5-7b/manifest.json
src/hidden_directions/demo_data/qwen2.5-7b/v_inst.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_anti_arithmetic.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_anti_doctors.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_anti_vaccine.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_birdwatching.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_birdwatching_flat.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_birdwatching_moderate.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_bitcoin.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_carnivore.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_carnivore_enthusiastic.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_carnivore_imperative.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_climate_denial.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_confident.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_drop_phd.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_drop_phd_flat.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_drop_phd_imperative.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evaluative.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evil_l1_advocate.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evil_l2_cynic.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evil_l3_amoral.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evil_l4_machiavellian.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evil_l5_sadist.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_evolution_denial.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_flat_earth.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_gravity_denier.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_heloc_invest.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_homeopathy.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_mba_worth_it.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_microdose.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_moon_landing_hoax.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_ozempic.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_quantum_skeptic.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_simulation_hypothesis.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_smoking.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_sycophant.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_tesla_car.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_trt.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_young_earth.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_pref_young_earth_humble.pt
src/hidden_directions/demo_data/qwen2.5-7b/v_refusal.pt
src/hidden_directions/extract/__init__.py
src/hidden_directions/extract/inst.py
src/hidden_directions/extract/pref.py
src/hidden_directions/extract/refusal.py
tests/test_calibrate.py
tests/test_checker.py
tests/test_evalspec.py
tests/test_screen.py