Research map / Large language models / Retrieval and knowledge
A Dense Subset Index for Collective Query Coverage
Kartik Nair, Pritish Chakraborty, Atharva Abhijit Tambat, Indradyumna Roy, Soumen Chakrabarti, Anirban Dasgupta, Abir De
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use DISCO instead of Exact Greedy for gold-set coverage | Exact Greedy · HotpotQA (|Sgold|=2) | Error(F): 0.69 -> 0.68 | Improved |
| use DISCO instead of Exact Greedy | Exact Greedy · HotpotQA (|Sgold|=2) | MAP: 0.83 -> 0.84 | No clear effect |
| use query-agnostic random pruning (Stochastic Greedy) instead of Exact Greedy | Exact Greedy · HotpotQA (|Sgold|=2) | MAP: 0.83 -> 0.19 | Hurt |
| use Lazier-than-lazy Greedy random pruning instead of Exact Greedy | Exact Greedy · HotpotQA (|Sgold|=2) | Error(F): 0.69 -> 3.09 | Hurt |
| use single-vector MUVERA instead of multivector PLAID | PLAID · HotpotQA (|Sgold|=2) | MAP: 0.81 -> 0.49 | Hurt |
| add early pooling instead of late pooling of projected queries (ablation) | DISCO · MS-Marco and Pooled | coverage and efficiency | Improved |
| increase number of random hyperplanes R in approximate marginal gain (ablation) | DISCO · NFCorpus and Writing | average coverage | Improved |
| use DISCO index instead of greedy variants | Exact Greedy · MS-Marco | efficiency: at least 100× | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PLAID — PLaID++: A Preference Aligned Language Model for Targeted Inorganic Materials Design (ICML 2026) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Distribution-Driven Dense Retrieval: Modeling Many-to-One Query-Document Relationship (AAAI 2025 · similarity 0.87)
- SoftMatcha: A Soft and Fast Pattern Matcher for Billion-Scale Corpus Searches (ICLR 2025 · similarity 0.86)
- FreshStack: Building Realistic Benchmarks for Evaluating Retrieval on Technical Documents (NeurIPS 2025 · similarity 0.85)
- Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layers (NeurIPS 2025 · similarity 0.84)
- Contextual Tokenization for Graph Inverted Indices (NeurIPS 2025 · similarity 0.84)
- POQD: Performance-Oriented Query Decomposer for Multi-vector retrieval (ICML 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
