מחקר מעמיק של דגמי שפה קטנים במשימות תפיסה רציונלית

תקציר מקורי באנגליתarXiv:2610.08680v1 Announce Type: new Abstract: Endpoint accuracy on abstract-reasoning benchmarks does not reveal whether a language model has acquired a transferable rule or fit distribution-specific regularities. We study this distinction in small language models on the ARC-TGI benchmark, which organizes abstract grid transformations into controllable task families and supports resampling, spatial shifts, and cross-benchmark transfer. Across more than 1,000 runs, we profile decoder-only, encoder--decoder, and mixture-of-experts model families under supervised fine-tuning. We examine the efficiency and stability of skill acquisition, robustness beyond the training distribution, interactions with model family and task formulation, and layer-wise attention signatures that accompany behavio