2026-10-26 ~ 11-01 · AI Search
Generative · Personalization · AI Search · Infra · Embodied AI
- 1. Semantically Rich, Metrically Affect-Blind: A Geometric Blind Spot in Modern Text Embedders · 📖 Deep dive
Search a vector index for what users dislike about a product, and the reviews you get back are usually on topic but a mix of praise and complaints. This paper measures why…
Appeared in: 10-26
- 2. LightRetriever: A LLM-based Text Retrieval Architecture with Extremely Faster Query Inference · 📖 Deep dive
In LLM-based text retrieval, documents can be encoded offline, but queries arrive in real time, so every query runs the full LLM, which keeps throughput low and serving costly…
Appeared in: 10-27
- 3. System One models Are Not Better Rerankers · 📖 Deep dive
Using the new decision models to rerank search results does not beat a purpose-tuned reranker, concludes this University of Passau paper. Decision models, also called System One models…
Appeared in: 10-27
- 4. EmbeddingGemma 2: Unified and Efficient Multimodal Representations · 📖 Deep dive
To search text, images, video and audio with one embedding model, the usual route is to transcribe audio and caption images first, then hand everything to a text model. That costs extra compute and throws away tone of…
Appeared in: 10-29
- 5. PersonalQuery-Benchmark:Evaluating Retrieval Robustness to Personalized User Expression Differences · 📖 Deep dive
Two shoppers want the same product. One asks for strawberry, lactose-free Greek yogurt in a short phrase; the other writes a longer sentence and misspells a word along the way…
Appeared in: 10-26
- 6. What the Query Remembers: Query Abstraction Governs Lexical and Dense Retrieval of Classical Arabic Poetry · 📖 Deep dive
When searching a classical Arabic poetry archive, which works better: keyword search with BM25 or an embedding model? This paper finds that the answer depends mainly on how much of the original wording the user still…
Appeared in: 10-26
- 7. Closing the Modality Gap via Simplex-Constrained Representations · 📖 Deep dive
In contrastive models like CLIP and CLAP, the two modalities are trained to align yet their embeddings often sit in offset regions, a phenomenon called the modality gap. This paper leaves the backbone untouched and…
Appeared in: 10-27
- 8. ACARE: Agent-Context-Aware Reranker for Iterative Search · 📖 Deep dive
Search agents often send sub-queries of just a few keywords, so a sub-query-only reranker promotes documents that match the words but do not help the real question. ACARe gives the reranker the original question and…
Appeared in: 10-28
- 9. CA-HNM: Constraint-Aware Hard Negative Mining for Dense Retrieval with Structured Domain Models · 📖 Deep dive
Dense retrievers usually take hard negatives from top-ranked documents that are not labeled positive, but ranking high only shows a document resembles the query, not why it should be excluded…
Appeared in: 10-28
- 10. Retriever-Free Retrieval: Navigating Documents with Decision Models · 📖 Deep dive
Most RAG systems still retrieve with BM25 or vector similarity, and evidence the retriever misses cannot be recovered by any generator. DecisionRAG drops the retriever and has a decision model browse like a person in a…
Appeared in: 10-28