19–21 Nov 2026
"Alexandru Ioan Cuza" University of Iasi and "Gheorghe Asachi" Technical University of Iași
Europe/Bucharest timezone

When Do Distilled Surrogates Preserve Algorithmic Amplification? A Simulation Study with LLM Agents

20 Nov 2026, 10:45
15m
Room 3

Room 3

Paper presentation Social Networking and Services Open Source Education and Research & SNS

Speaker

Teodor Milea (UNSTPB)

Description

Auditing algorithmic amplification on very largeonline platforms, as envisioned under Article 40 of the EU DigitalServices Act, is obstructed by two constraints: auditors rarelyhave white-box access to production rankers, and counterfactualinterventions cannot be run on a live platform. A naturalworkaround is to distill a surrogate of the ranker and studyit inside an agent-based simulation, but this is only soundif conclusions transfer from surrogate to target. We test thistransfer in OASIS, an LLM-agent simulation of X/Twitter whoseranker is TwHIN-BERT. We distill two surrogates under differentaccess assumptions, one reusing the target’s first four transformerlayers and one training a compact sentence encoder on thetarget’s embeddings, and run matched simulations in whichonly the ranker differs. When a single post spreads throughthe population of its real cascade, all rankers reproduce thescale of Twitter15/16 cascades equally well. When 100 postscompete for two recommendation slots, the surrogates share only9–29% of their first recommendations with their teacher, andoutcomes such as the share of reposts on false rumors varyacross rankers. A minor implementation change in the targetitself, its position-embedding handling, reduces agreement withthe deployed ranker to 7%. Embedding-level fidelity thereforedoes not guarantee fidelity of exposure outcomes, and single-postcascades are a weak test of a ranker. These preliminary, single-seed results motivate validating surrogate-based audits in settingswhere the ranker actually filters content.

Author

Teodor Milea (UNSTPB)

Co-authors

Ms Alexandra Neacșu (UNSTPB) Rodica Siminiuc (Technical University of Moldova) Prof. Răzvan Victor RUGHINIȘ (National University of Science and Technology POLITEHNICA Bucharest)

Presentation materials

There are no materials yet.