AI
Robust Is Salient: An Informed Adversary Moves the Optimal Signal onto the Salience Pole
Publication: cs.AI updates on arXiv.org
Author: Cris Huynh
Source published: 02/10/2026, 04:00:00 UTC
Briefing created: 02/10/2026, 04:30:11 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 02/10/2026, 04:03:25 UTC.
Briefings
An informed adversary in a constrained signalling channel optimally selects the signal that best describes the target rather than the one maximizing posterior probability. On 108 confirmatory items, this adversary-robust optimum aligns exactly with the salience pole, differing from the salience-to-Bayes coordinate only on 2,748 items. The study introduces an adversary to a forced-choice task, showing that as persuasion budget increases, the optimal signal shifts from posterior-maximizing to margin-maximizing. At zero budget, the game reproduces the original oracle model with a listener temperature of 1.182% of the pool exhibits this shift, and each item's critical budget is exact.
Key points
- On 108 confirmatory items, the adversary-robust optimum aligns exactly with the salience pole, differing from the salience-to-Bayes coordinate only on 2,748 items.
- As persuasion budget increases, the optimal signal shifts from the posterior-maximizing option to the margin-maximizing one, with the critical budget being exact for each item.
Context and limitations
The study is limited to 108 confirmatory items and a 200,000-item pool; the diagnostic check is cheap and cannot determine whether the movement is toward the adversary-aware optimum or toward salience because the two options are identical.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 02/10/2026, 04:30:11 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeifjgjueu5nhzuazwtowa7e5ewykzoybig2uco5phgpme2udw2rafq