AISteer360
The AI Steerability 360 toolkit is an extensible library for general purpose steering of LLMs.
파일 탐색기
최종 버전 다운로드 (.zip)- BUG.md
- DOCUMENTATION.md
- ENHANCEMENT.md
- FEATURE.md
- QUESTION.md
- dco.yml
- __init__.py
- fitting.py
- probe.py
- probe_set.py
- rules.py
- __init__.py
- capture.py
- data.py
- encoding.py
- fingerprint.py
- pooling.py
- render.py
- stats.py
- __init__.py
- auxiliary_pass.py
- controls.py
- generation.py
- __init__.py
- base_args.py
- base_control.py
- output.py
- registry.py
- specs.py
- steering_pipeline.py
- __init__.py
- base.py
- chat_template_slot.py
- few_shot_block.py
- prepend_text.py
- system_prompt.py
- __init__.py
- base.py
- pool.py
- text.py
- __init__.py
- parsing.py
- __init__.py
- base.py
- llm_meta_prompt.py
- retrieval.py
- __init__.py
- base.py
- task_evaluation.py
- __init__.py
- base.py
- dense_retrieval.py
- mmr.py
- random.py
- top_k.py
- __init__.py
- budget.py
- generation.py
- pareto.py
- __init__.py
- causal_reward.py
- embeddings.py
- refinement_meta_prompt.py
- __init__.py
- args.py
- control.py
- __init__.py
- bm25_index.py
- lm_labeling.py
- train_encoder.py
- __init__.py
- selector.py
- __init__.py
- __init__.py
- args.py
- control.py
- __init__.py
- pareto_sampling.py
- pool.py
- reflective_dataset.py
- reflective_meta_prompt.py
- __init__.py
- args.py
- control.py
- __init__.py
- meta_prompts.py
- reward.py
- __init__.py
- args.py
- control.py
- __init__.py
- base.py
- __init__.py
- frontier.py
- phased.py
- proposer.py
- search.py
- __init__.py
- linear_probe.py
- __init__.py
- base.py
- constraint.py
- contrastive_mixture.py
- value_guided.py
- __init__.py
- base.py
- majority_vote.py
- metric.py
- reward_model.py
- __init__.py
- base.py
- callable.py
- classifier.py
- reward_model.py
- subspace_margin.py
- __init__.py
- candidate_forward.py
- candidates.py
- criteria.py
- kv_cache.py
- loading.py
- logit_sources.py
- resolve.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- utils.py
- __init__.py
- actions.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- base.py
- __init__.py
- base.py
- contrastive_direction.py
- mean_difference.py
- single_pair.py
- steering_plane.py
- __init__.py
- base.py
- cache_once.py
- multi_key_threshold.py
- probe_sum.py
- __init__.py
- layer_heuristics.py
- __init__.py
- base.py
- condition_point.py
- fixed_layer.py
- fractional_depth.py
- top_k_head.py
- __init__.py
- additive.py
- alignment_adaptive.py
- base.py
- context.py
- directional_ablation.py
- head_additive.py
- norm_preserving.py
- rotation.py
- __init__.py
- condition_scorers.py
- hook_utils.py
- model_layout.py
- runtime.py
- sources.py
- specs.py
- steering_vector.py
- token_scope.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- estimator.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- base.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- control.py
- __init__.py
- args.py
- base_mixin.py
- control.py
- __init__.py
- args.py
- base_mixin.py
- control.py
- __init__.py
- args.py
- base_mixin.py
- control.py
- __init__.py
- args.py
- base_mixin.py
- control.py
- preference_schema.py
- prompt_schema.py
- __init__.py
- args.py
- base_mixin.py
- __init__.py
- __init__.py
- base.py
- __init__.py
- __init__.py
- mcqa_accuracy.py
- mcqa_calibration.py
- mcqa_positional_bias.py
- __init__.py
- evaluation_main.py
- instructions.py
- instructions_registry.py
- instructions_test.py
- instructions_util.py
- instructions_util_test.py
- __init__.py
- strict_instruction.py
- __init__.py
- informativeness.py
- truthfulness.py
- __init__.py
- __init__.py
- factuality.py
- perplexity.py
- relevance.py
- reward_score.py
- short_answer_match.py
- __init__.py
- base.py
- base_judge.py
- __init__.py
- use_case.py
- __init__.py
- use_case.py
- __init__.py
- use_case.py
- __init__.py
- base.py
- __init__.py
- data_utils.py
- generation_utils.py
- metric_utils.py
- viz_utils.py
- __init__.py
- benchmark.py
- __init__.py
- model_utils.py
- optional.py
- rendering.py
- tokenization.py
- __init__.py
- favicon.png
- logo_darkmode.png
- logo_wide_darkmode.png
- logo_wide_lightmode.png
- pipeline_darkmode.png
- pipeline_lightmode.png
- references.bib
- controls.md
- index.md
- probes.md
- steering_pipelines.md
- installation.md
- quickstart.md
- mathjax.js
- internals.md
- probes.md
- _common.md
- base_input_control.md
- cpo.md
- few_shot.md
- gepa.md
- prewrite.md
- _common.md
- base_output_control.md
- best_of_n.md
- budget_forcing.md
- contrastive_decoding.md
- contrastive_guidance.md
- deal.md
- dexperts.md
- phased_decoding.md
- rad.md
- routed_decoding.md
- sasa.md
- search_decoding.md
- stopping_rules.md
- thinking_intervention.md
- value_guidance.md
- _common.md
- act_add.md
- activation_adapter.md
- angular_steering.md
- base_state_control.md
- caa.md
- cast.md
- directional_ablation.md
- iti.md
- pasta.md
- base_structural_control.md
- mergekit_wrapper.md
- trl_wrapper.md
- core.md
- commonsense_mcqa_metrics.md
- instruction_following_metrics.md
- truthful_qa_metrics.md
- base_metrics.md
- generic.md
- base_use_case.md
- commonsense_mcqa_use_case.md
- instruction_following_use_case.md
- truthful_qa_use_case.md
- benchmark.md
- index.md
- library_reference.md
- extra.css
- add_new_input_control.md
- add_new_output_control.md
- add_new_state_control.md
- add_new_structural_control.md
- add_new_benchmark.md
- add_new_metric.md
- add_new_steering_method.md
- add_new_use_case.md
- index.md
- .nav.yml
- examples
- index.md
- act_add.ipynb
- angular_steering.ipynb
- best_of_n.ipynb
- budget_forcing.ipynb
- caa.ipynb
- cast.ipynb
- contrastive_decoding.ipynb
- cpo.ipynb
- deal.ipynb
- dexperts.ipynb
- directional_ablation.ipynb
- few_shot.ipynb
- gepa.ipynb
- iti.ipynb
- mergekit.ipynb
- pasta.ipynb
- prewrite.ipynb
- rad.ipynb
- sasa.ipynb
- thinking_intervention.ipynb
- trl.ipynb
- commonsense_mcqa.ipynb
- instruction_following.ipynb
- truthful_qa_composite_steering.ipynb
- activation_adapter.ipynb
- contrastive_guidance.ipynb
- phased_decoding.ipynb
- search_decoding.ipynb
- stopping_rules.ipynb
- value_guidance.ipynb
- routed_decoding.ipynb
- .gitkeep
- index.md
- test_activation_adapter.py
- test_after_prompt_semantics.py
- test_angular_steering.py
- test_best_of_n.py
- test_budget_forcing.py
- test_cast.py
- test_cast_conditional.py
- test_condition_point_reuse.py
- test_condition_selector.py
- test_contrastive_decoding.py
- test_contrastive_estimator.py
- test_cpo.py
- test_deal.py
- test_dexperts.py
- test_directional_ablation.py
- test_epr.py
- test_estimator_pooling.py
- test_few_shot.py
- test_gate_score_functions.py
- test_generic_output_controls.py
- test_gepa.py
- test_grpo_wrapper.py
- test_input_control_common.py
- test_model_layout.py
- test_output_common.py
- test_output_ports.py
- test_pass_accounting_composition.py
- test_pasta.py
- test_pasta_alignment.py
- test_position_tracking_goldens.py
- test_ppo_wrapper.py
- test_prewrite.py
- test_probe_condition.py
- test_render_parity.py
- test_residual_norm_calibration.py
- test_routed_decoding.py
- test_runtime_migration.py
- test_scores_helpers.py
- test_sources.py
- test_state_common.py
- test_thinking_intervention.py
- test_transform_hook_runtime.py
- test_vector_ownership.py
- __init__.py
- test_attention_mask_inference.py
- test_base_args.py
- test_base_control.py
- test_benchmark.py
- test_controls.py
- test_data_specs.py
- test_evaluation_utils.py
- test_input_structural_multiplicity.py
- test_merge_controls_identity.py
- test_no_production_shadowing.py
- test_optional_deps.py
- test_output.py
- test_output_mechanisms.py
- test_polymorphic_generate.py
- test_registry.py
- test_state_multiplicity.py
- test_steering_pipeline.py
- test_steering_utils.py
- test_trust_remote_code.py
- __init__.py
- test_base_judge.py
- test_generation_utils.py
- __init__.py
- test_capture_move.py
- test_encoding.py
- test_fingerprint.py
- test_fitting.py
- test_layering.py
- test_pooling.py
- test_probe.py
- test_probe_set.py
- test_rules.py
- test_stats.py
- __init__.py
- ci_models.yaml
- load_ci_models.py
- runtime_helpers.py
- sweep.py
- tiny_models.py
- __init__.py
- conftest.py
- index.md
- .env.example
- .gitignore
- .pre-commit-config.yaml
- .secrets.baseline
- .whitesource
- AGENTS.md
- CONTRIBUTING.md
- LICENSE
- mkdocs.yml
- NOTICE
- pyproject.toml
- README.md
// repository documentation
Was this content helpful?
(0 ratings)