AI & Machine Learning · Thu, Oct 8, 2026

Machine learning, AI, natural language processing, computer vision and multi-agent systems.

RSS feed
arXiv cs.LG

How to train your model organism

In briefTraining model organisms with multi-objective validation for interpretability research

SummaryThe paper argues that training model organisms, models with installed behaviors like backdoors or sycophancy, to a single objective is insufficient for evaluating interpretability methods. It proposes validating them on three objectives: target-behavior installation, general-capability preservation, and output naturalness. The authors show these validation metrics predict how well interpretability methods recover the installed behavior, and they introduce a model-merging-based multi-objective training approach.

Xilin Wang, David Bau, Byron C. Wallace

arXiv cs.AI · cs.CL · cs.CV

GAGR-Lab: Evaluating Joint Spatial-Geometric and Analytic Function Reasoning

In briefGAGR-Lab: A Framework for Testing Joint Spatial-Geometric and Analytic Function Reasoning

SummaryThe paper presents GAGR-Lab, a framework that measures how well a model turns a spatial scene into a symbolic function whose executed curve meets geometric constraints, using Cartesian game scenes and a Rust trajectory executor. In a bounded pilot, the hosted model Llama 3.2 11B Vision Instruct produced 432 attempts with no target hits, while a privileged analytic search control succeeded on 600 directional cases. The authors outline a staged prospective study plan, and the full difficulty matrix and comparative model results remain untested.

Jingyao Zhang, Yun Li, Lu Han

arXiv cs.LG

Robust Decentralized Fairness Auditing

In briefRobust Decentralized Fairness Auditing (Auditopus)

SummaryThis paper proposes Auditopus, a method for auditing the fairness of large language models in which several auditors query the model separately and share only cumulative statistics vectors, not raw queries. The authors show that a single adversarial auditor can fabricate these vectors to make an unfair model appear fair. To counter this, honest auditors down-weight any auditor whose statistics are inconsistent with earlier rounds. Experiments report reduced audit error compared with no defense and standard robust aggregation, and state that unfair models are not passed as fair even when 49% of auditors are adversarial.

Sayan Biswas, Jade Garcia Bourr\'ee, Anne-Marie Kermarrec et al.