Add availability heuristic bias to CoBRA pipeline#3
Open
tbhat-ops wants to merge 1 commit intoAISmithLab:masterfrom
Open
Add availability heuristic bias to CoBRA pipeline#3tbhat-ops wants to merge 1 commit intoAISmithLab:masterfrom
tbhat-ops wants to merge 1 commit intoAISmithLab:masterfrom
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Paradigms included:
Files Created (3)
data/availability/availability_judgment_of_word_frequency.json
24 scenarios testing availability bias via letter/prefix position frequency judgment across 3 sources (English word, city name, last name) × 8 units.
data/availability/availability_fame_frequency_and_recall.json
15 scenarios testing availability bias via name-list frequency judgment, each with ~39 mixed-order names split into salient/less-salient groups across 5 domains.
examples/availability/utils_availability.py
Three functions: load_availability_scenarios() to load seed data with scenario-type dispatch, create_availability_dataset() for hand-crafted RepE training, create_availability_dataset_from_generated() for generated-data RepE training.
Files Modified (6)
examples/unified_bias/utils_bias.py (6 edits)
Registered availability in the central registry: added availability_dir to sys.path, imported utils_availability functions, added config dict entry with both paradigms, added dispatcher in load_experiment_scenarios(), added 'availability' to get_all_bias_types(), added MCQ options + templates for both paradigms in get_mcq_options_and_templates().
control/prompt_experiment.py (1 edit)
Added availability entry to bias_descriptions dict in _generate_likert_control_prompt() with terse descriptions for 0% (statistical) vs 100% (recency-biased) endpoints.
control/api_prompt_experiment.py (1 edit)
Added availability entry to bias_descriptions dict in _generate_likert_control_prompt() with verbose descriptions matching API-layer style.
examples/unified_bias/run_pipelines.py (1 edit)
Updated --bias argument help text to include 'availability' in the list of allowed bias types.
generator/scenario_generator.py (3 edits)
Added "availability" key to self.bias_prompts dict, added _get_availability_generation_prompt() method with both paradigm examples and generation instructions, updated argparse choices and default lists and generate_full_dataset() default to include 'availability'.
.gitignore updates to avoid committing plots generated in test run