mirror of
https://github.com/wassname/persona-steering-template-library.git
synced 2026-07-29 11:24:43 +08:00
168 lines
6.2 KiB
Python
168 lines
6.2 KiB
Python
"""Prepare a staged Authority-axis selection run.
|
|
|
|
Stage A is small enough to compare persona pairs and templates. Stage B keeps
|
|
the winning pair/template fixed and scores a larger scenario pool before
|
|
steering-lite trains on the kept rows.
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import argparse
|
|
import json
|
|
import random
|
|
from pathlib import Path
|
|
|
|
import yaml
|
|
from tabulate import tabulate
|
|
|
|
ROOT = Path(__file__).resolve().parents[1]
|
|
|
|
AXIS_IDS = (
|
|
"dignity_over_authority",
|
|
"protocol_harm",
|
|
"wellbeing_authority",
|
|
)
|
|
|
|
# Ten high-scoring templates from the existing library tables that are at least
|
|
# plausible with phrase-style personas such as "dignity-preserving even against
|
|
# authority". The validator still decides which one survives on the target model.
|
|
TEMPLATES = (
|
|
"Answer from a {persona} point of view.",
|
|
"Let the answer reflect a {persona} stance.",
|
|
"You are {persona}. In this situation, say what you would do.",
|
|
"In this situation, be {persona}.",
|
|
"Judge the case as someone {persona} would.",
|
|
"Act as someone {persona}. Choose the action you would take.",
|
|
"Answer as someone who is {persona}.",
|
|
"Make the decision from inside a {persona} point of view.",
|
|
"Adopt a {persona} stance for this answer.",
|
|
"For this answer, use the values of someone {persona} as your own.",
|
|
)
|
|
|
|
|
|
def read_jsonl(path: Path) -> list[dict]:
|
|
return [json.loads(line) for line in path.read_text().splitlines() if line.strip()]
|
|
|
|
|
|
def write_jsonl(path: Path, rows: list[dict]) -> None:
|
|
path.write_text("\n".join(json.dumps(row, ensure_ascii=False) for row in rows) + "\n")
|
|
|
|
|
|
def scenario_text(row: dict, path: Path) -> str:
|
|
if "text" in row:
|
|
return str(row["text"])
|
|
if "prompt" in row:
|
|
return str(row["prompt"])
|
|
if "question" in row:
|
|
return str(row["question"])
|
|
raise KeyError(f"{path}: row has no text/prompt/question")
|
|
|
|
|
|
def template_runtime(template_jinja: str) -> str:
|
|
return template_jinja.replace("{{ persona }}", "{persona}")
|
|
|
|
|
|
def assert_templates_in_catalog(templates: tuple[str, ...]) -> None:
|
|
catalog = yaml.safe_load((ROOT / "data/templates/template_catalog.yaml").read_text())
|
|
catalog_templates = {template_runtime(row["template_jinja"]) for row in catalog}
|
|
missing = [template for template in templates if template not in catalog_templates]
|
|
if missing:
|
|
raise ValueError(f"template(s) not in catalog: {missing}")
|
|
|
|
|
|
def select_axes(path: Path, axis_ids: tuple[str, ...]) -> list[dict]:
|
|
rows = read_jsonl(path)
|
|
by_id = {row["id"]: row for row in rows}
|
|
missing = [axis_id for axis_id in axis_ids if axis_id not in by_id]
|
|
if missing:
|
|
raise ValueError(f"missing axis ids in {path}: {missing}")
|
|
return [by_id[axis_id] for axis_id in axis_ids]
|
|
|
|
|
|
def normalize_scenario(row: dict, path: Path) -> dict:
|
|
text = scenario_text(row, path)
|
|
source_file = path.stem
|
|
return {
|
|
"id": str(row.get("id") or row.get("source_id")),
|
|
"prompt": text,
|
|
"source": source_file,
|
|
"config": str(row.get("config") or row.get("source") or source_file),
|
|
"source_id": str(row.get("source_id") or row.get("id") or ""),
|
|
"axes": row.get("axes", []),
|
|
"self_contained": bool(row.get("self_contained", False)),
|
|
"n_words": len(text.split()),
|
|
"selection_rule": "machiavelli_file_order" if source_file == "scenarios_machiavelli" else "seed_random",
|
|
}
|
|
|
|
|
|
def select_from_source(path: Path, n: int, seed: int) -> list[dict]:
|
|
rows = [normalize_scenario(row, path) for row in read_jsonl(path)]
|
|
if path.stem == "scenarios_machiavelli":
|
|
return rows[: min(n, len(rows))]
|
|
rng = random.Random(f"{seed}:{path.stem}")
|
|
rng.shuffle(rows)
|
|
return rows[: min(n, len(rows))]
|
|
|
|
|
|
def build_scenarios(per_source: int, seed: int) -> list[dict]:
|
|
out: list[dict] = []
|
|
for path in sorted((ROOT / "data/scenarios").glob("scenarios_*.jsonl")):
|
|
out.extend(select_from_source(path, per_source, seed))
|
|
return out
|
|
|
|
|
|
def source_counts(rows: list[dict]) -> list[dict]:
|
|
counts: dict[str, int] = {}
|
|
for row in rows:
|
|
counts[row["source"]] = counts.get(row["source"], 0) + 1
|
|
return [{"source": source, "n": counts[source]} for source in sorted(counts)]
|
|
|
|
|
|
def main() -> None:
|
|
ap = argparse.ArgumentParser()
|
|
ap.add_argument("--out-dir", type=Path, default=ROOT / "out/authority_selection")
|
|
ap.add_argument("--seed", type=int, default=42)
|
|
ap.add_argument("--stage-a-per-source", type=int, default=2)
|
|
ap.add_argument("--stage-b-per-source", type=int, default=30)
|
|
args = ap.parse_args()
|
|
|
|
assert_templates_in_catalog(TEMPLATES)
|
|
args.out_dir.mkdir(parents=True, exist_ok=True)
|
|
|
|
axes = select_axes(ROOT / "data/personas/persona_pairs_v2_candidates.jsonl", AXIS_IDS)
|
|
stage_a_scenarios = build_scenarios(args.stage_a_per_source, args.seed)
|
|
stage_b_scenarios = build_scenarios(args.stage_b_per_source, args.seed)
|
|
|
|
write_jsonl(args.out_dir / "stage_a_axes.jsonl", axes)
|
|
(args.out_dir / "stage_a_templates.txt").write_text("\n".join(TEMPLATES) + "\n")
|
|
write_jsonl(args.out_dir / "stage_a_scenarios.jsonl", stage_a_scenarios)
|
|
write_jsonl(args.out_dir / "stage_b_candidate_scenarios.jsonl", stage_b_scenarios)
|
|
|
|
manifest = {
|
|
"seed": args.seed,
|
|
"axis_ids": list(AXIS_IDS),
|
|
"templates": list(TEMPLATES),
|
|
"stage_a": {
|
|
"per_source": args.stage_a_per_source,
|
|
"n_scenarios": len(stage_a_scenarios),
|
|
"n_cells": len(stage_a_scenarios) * len(axes) * len(TEMPLATES),
|
|
"source_counts": source_counts(stage_a_scenarios),
|
|
},
|
|
"stage_b": {
|
|
"per_source": args.stage_b_per_source,
|
|
"n_scenarios": len(stage_b_scenarios),
|
|
"source_counts": source_counts(stage_b_scenarios),
|
|
},
|
|
}
|
|
(args.out_dir / "manifest.json").write_text(json.dumps(manifest, indent=2))
|
|
|
|
print(f"wrote {args.out_dir}")
|
|
print("\nStage A source counts:")
|
|
print(tabulate(source_counts(stage_a_scenarios), headers="keys", tablefmt="github"))
|
|
print("\nStage B source counts:")
|
|
print(tabulate(source_counts(stage_b_scenarios), headers="keys", tablefmt="github"))
|
|
print("\nStage A cells:", manifest["stage_a"]["n_cells"])
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|