{"ID":3004929,"CreatedAt":"2026-06-03T03:09:48.883664427Z","UpdatedAt":"2026-06-04T19:14:31.964469513Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.03091","arxiv_id":"2606.03091","title":"BAHSD: Bridging the Long-tail Gap via Adaptive Distillation in Black-box Sequential Recommendation","abstract":"Sequential recommendation systems are widely adopted but often deployed as black-box APIs, which has driven recent interest in model extraction to replicate their capabilities locally. However, the long-tail distribution induces severe signal heterogeneity: dense head sequences trigger the solidification of teacher preference, biasing extraction toward local patterns, while sparse tail sequences yield flat, noisy predictions. Existing one-size-fits-all extraction overlooks this disparity, resulting in noise overfitting and suboptimal knowledge transfer. We propose BAHSD, a black-box adaptive distillation framework that handles signal heterogeneity via a multi-scale consistency probing mechanism to implicitly quantify signal reliability. Based on this, an adaptive hierarchical objective is designed: dynamic-temperature KL divergence mitigates preference solidification for high-confidence signals, while ranking consistency and InfoNCE contrastive learning provide noise-robust enhancement for low-confidence signals. BAHSD consistently outperforms baselines, achieving up to 4.98\\% gain over the teacher and 80\\%+ improvement on tail users, offering a plug-and-play solution for high-fidelity black-box recommendation extraction.","short_abstract":"Sequential recommendation systems are widely adopted but often deployed as black-box APIs, which has driven recent interest in model extraction to replicate their capabilities locally. However, the long-tail distribution induces severe signal heterogeneity: dense head sequences trigger the solidification of teacher pre...","url_abs":"https://arxiv.org/abs/2606.03091","url_pdf":"https://arxiv.org/pdf/2606.03091v1","authors":"[\"Xi Zhou\",\"Famin Wu\",\"Mingming Li\",\"Hongyue Zhang\",\"Jiao Dai\",\"Jizhong Han\",\"Tao Guo\"]","published":"2026-06-02T03:26:25Z","proceeding":"cs.IR","tasks":"[\"cs.IR\",\"cs.AI\"]","methods":"[]","has_code":false}