Allow negative utility factors for "misère" search#31
Draft
ChinChangYang wants to merge 1 commit into
Draft
Conversation
ChinChangYang
changed the title
Relax the lower bound of winLossUtilityFactor, staticScoreUtilityFactor, and dynamicScoreUtilityFactor from 0.0 to -1.0 in config parsing. Negative factors invert the search objective so both players search for the most-losing move, which models the sub-20k regime (where weak opponents play worse than random) better than a normal win-seeking search. Strength is then calibrated via temperature. Add gtp_human_misere_example.cfg demonstrating the setup, with the necessary guardrails (useUncertainty=false to avoid a NaN from the negative-factor uncertainty term, shallow search since the value net is unreliable for deep losing lines, and human-SL exploration to keep blunders human-shaped). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0124xb2Xkir35BQHpuxthSPS
ChinChangYang
force-pushed
the
claude/chosenmovetemperatureonlybelowprob-epb1li
branch
from
eb5b2c0 to
ce2172e
Compare
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Relax the lower bound of winLossUtilityFactor, staticScoreUtilityFactor,
and dynamicScoreUtilityFactor from 0.0 to -1.0 in config parsing. Negative
factors invert the search objective so both players search for the
most-losing move, which models the sub-20k regime (where weak opponents
play worse than random) better than a normal win-seeking search. Strength
is then calibrated via temperature.
Add gtp_human_misere_example.cfg demonstrating the setup, with the
necessary guardrails (useUncertainty=false to avoid a NaN from the
negative-factor uncertainty term, shallow search since the value net is
unreliable for deep losing lines, and human-SL exploration to keep
blunders human-shaped).
Co-Authored-By: Claude Opus 4.8 noreply@anthropic.com
Claude-Session: https://claude.ai/code/session_0124xb2Xkir35BQHpuxthSPS