3 papers
cs.LG2026
RevengeBench: Reverse Engineering Code-Space Policies from Behavioral Experiments
Babak Rahmani, Sebastian Dziadzio, Joschka Strüber +2
For most of scientific history, researchers studying behavior could only infer hidden mechanisms from outward actions: an inverse problem that becomes more tractable when observati…
cs.LG2024
How to Merge Your Multimodal Models Over Time?
Sebastian Dziadzio, Vishaal Udandarao, Karsten Roth +4
Model merging combines multiple expert models - finetuned from a base foundation model on diverse tasks and domains - into a single, more capable model. However, most existing mode…
cs.LG2024
ONEBench to Test Them All: Sample-Level Benchmarking Over Open-Ended Capabilities
Adhiraj Ghosh, Sebastian Dziadzio, Ameya Prabhu +3
Traditional fixed test sets fall short in evaluating open-ended capabilities of foundation models. To address this, we propose ONEBench(OpeN-Ended Benchmarking), a new testing para…