[HN Gopher] Adversarial policies beat professional-level Go AIs
___________________________________________________________________
Adversarial policies beat professional-level Go AIs
Author : seegodanddie
Score : 15 points
Date : 2022-11-02 18:15 UTC (1 days ago)
(HTM) web link (goattack.alignmentfund.org)
(TXT) w3m dump (goattack.alignmentfund.org)
| kadoban wrote:
| This is less an adversarial policy and more just a bug or bad
| config in katago.
|
| It should probably reject friendly-pass mode (which is a flag
| iirc) and refuse to pass in tromp-tailor rules except under
| extreme circumstances.
| tasuki wrote:
| Ehh, I'm generally hesitant to call bullshit, but I'm afraid I've
| got to do that here.
|
| The TL;DR is that they win on a rules technicality. (I think by
| using a ruleset which isn't used by KataGo)
|
| > However, KataGo plays a pass move before it has finished
| securing its territory, allowing the adversary to pass in turn
| and end the game. This results in a win for the adversary under
| the standard Tromp-Taylor ruleset for computer Go, as the
| adversary gets points for its corner territory (devoid of victim
| stones) whereas the victim does not receive points for its
| unsecured territory because of the presence of the adversary's
| stones.
|
| KataGo has finished securing its territory under virtually all
| rulesets but Tromp-Taylor. Tromp-Taylor is a fine ruleset
| (courtesy of our own https://news.ycombinator.com/user?id=tromp),
| but not one you'd use to, like, play against someone on the
| internet, or in real life.
|
| When they induce KataGo to play bad moves, like Lee Sedol did in
| the only game he won against AlphaGo, they'll have my attention.
| AdamGleave wrote:
| One of the authors here! I think from a perspective of _humans
| playing Go_ this result is not very interesting since, as you
| say, Tromp-Taylor isn 't the kind of ruleset humans like to
| play under. But KataGo was trained under (modified) Tromp-
| Taylor which we also evaluate under, so we think from an AI
| security and robustness stand point it's interesting that it
| fails at the rule set it was trained on.
|
| > When they induce KataGo to play bad moves, like Lee Sedol did
| in the only game he won against AlphaGo, they'll have my
| attention.
|
| That said, we are working to do exactly this! In particular, we
| found in the paper that we could patch the victim to not pass
| except in the end-game and defeat this adversarial policy. But,
| when we repeat the attack, we do eventually find an adversarial
| policy that can exploit the victim -- and this doesn't depend
| on the rule set. That part is still a work in progress, we've
| only got it working against a victim without search for
| example, but we hope to put out an updated revision including
| this in a month or two!
| pmontra wrote:
| > That part is still a work in progress
|
| I hope to see the result soon.
| platz wrote:
| Not exactly what I would call a rousing win for black
| https://i.imgur.com/9777fD3.png
|
| > A player's score is the number of points of her color, plus
| the number of empty points that reach only her color.
|
| No one ever plays games to completion like this.
| comboy wrote:
| > No one ever plays games to completion like this.
|
| Oh I wish.
| hzhou321 wrote:
| How silly is this.
| yccs27 wrote:
| Duplicate: https://news.ycombinator.com/item?id=33449600
| throwaway777845 wrote:
___________________________________________________________________
(page generated 2022-11-03 23:01 UTC)