gSAP: differentiable loss optimizes global ranking across queries, replacing InfoNCE
serrjoa · x · 2026-10-08
A new arXiv paper introduces gSAP, a differentiable surrogate of Global Average Precision (gAP).
- Standard losses like InfoNCE and per-query AP surrogates rank per query and ignore cross-query comparability of similarity scores, which any single-threshold retrieval system needs;
- gSAP ranks all query-candidate pairs in one list, takes the same inputs as existing losses (similarity matrix + positive-pair matrix), and is a drop-in replacement agnostic to encoder, modality, and supervision source;
- By jointly considering all pairwise comparisons in a batch, it stays trainable at low temperatures where per-query surrogates run out of gradient;
- Swapping it into established recipes improves supervised metric learning, cross-modal alignment, and self-supervised pretraining—reportedly the first ranking loss to replace InfoNCE in the latter two.
More from Research
- Talk: how to RL-train an agent running inside a harness you didn't write — SergioPaniego · 2026-10-08
- AI labs accused of press-release deception: none of 520 claimed proofs Lean-formalized — gerardsans · 2026-10-08
- Commercial detector re-runs NeurIPS AI-text check: 7.3% of 2025 papers flagged vs Pangram's 1% — AltruisticCouple3491 · 2026-10-08
- Burned 300B Tokens with Nothing; Internal Model Broke Through in 3 Hours — burny_tech · 2026-10-08
- NVIDIA's UNREAL paper lets one LLM both retrieve and answer, lifting recall from 49% to 73% — mark_k · 2026-10-08
- NAMVIS: next-scale autoregression beats diffusion for multi-view synthesis, 3x faster — Ramil Khafizov · 2026-10-08