kaal:claim:7598518-026

Claim. The reputation update is an asymmetric multiplicative rule in which reputation is multiplied by one plus alpha when a vote matches the revealed key and by one minus gamma when it does not, with alpha fixed at 0.05, gamma at 0.10, and a floor applied after each update.

Type. design · Support. stated

Holds when

• The parent paper requires only that gamma exceed alpha; this design fixes both rates.
• Any single weight above one half is capped there and the excess redistributed among the other seated validators.

Source quote

The update is equation (3) of the RLRF paper, an asymmetric multiplicative rule. On tag t, reputation updates as ri,t ← ri,t(1 + α) when the vote matches the revealed key and ri,t ← ri,t(1 − γ) when it does not. The RLRF paper requires only (0) γ > α, and this design fixes α = 0. 05 and γ = 0. 10. Each reputation starts at ri,t = 1, and after each update the floor ri,t ≥ 0. 01 is applied.

From. Wulf A. Kaal, Can Reputation Prevent Model Collapse? A Preregistered Test of Reputation-Weighted Data Filtering (2026). Full text.

Cite as. Wulf A. Kaal, Can Reputation Prevent Model Collapse? A Preregistered Test of Reputation-Weighted Data Filtering (2026). SSRN: https://ssrn.com/abstract=7598518

Verify. sha256 of source PDF cd22d148b2b55c2e27f087dcaa61cdf44ad7022b400b355c31a9cd7e43a31749 at https://wulfkaal.com/papers/7598518.pdf

Content hash. 930f1b62d390c92b4d6dd19c2f5ca1abdd9963c39a393a748df4aa16e44db5d9