Weekly IR papers Vol.179: fixed-budget late interaction, freezing semantic IDs, search agents
_reachsumit · x · 2026-10-26
Sumit published Vol.179 of "Top Information Retrieval Papers of the Week" (Oct 19–25) covering 11 retrieval/recommendation papers:
- FLINT (Handa et al.): fixed-budget late-interaction retriever using a fixed number of latent vectors per query/document via Perceiver-style cross-attention pooling, addressing ColBERT's token-per-vector index growth
- Alibaba: freezing Semantic IDs for generative retrieval under item churn
- Perplexity AI: representation distillation for multimodal late-interaction retrieval
- Meta: training search agents to write their own retrieval pipelines
- Lee et al.: self-assessment as a training signal for search agents
- Plus unifying dense and late-interaction retrieval, item-level adaptive depth for sequential recommendation, long-tail item recovery without retraining, and explicit memory vs raw-history retrieval for long-term conversational agents
More from Research
- HIT Shenzhen unveils STILL: content-aware intra-layer hybrid attention to linearize LLMs — jiqizhixin · 2026-10-26
- Bambi quietly gains nonlinear model support, with a blog post showing how — twiecki · 2026-10-26
- Fudan and Tencent unveil DraMaze, a multi-agent pipeline for diverse microdrama scripts — Siyang He · 2026-10-26
- Alibaba's MultiAgent-OPD trains hierarchical multi-agent systems end-to-end with on-policy distillation — TaoLiveAIGC · 2026-10-26
- LoCA: two-stage pipeline adds multi-modality to geospatial location encoders — Panagiotis Koromilas · 2026-10-26
- Skipping eval checks means you're likely fitting benchmark biases and noise — antoine_chaffin · 2026-10-26