LoHoSearch is a new benchmark of 544 KG-constructed questions across 11 domains where the strongest search agent scores 34.74% and context strategies add at most 6.8%.
Fact, fetch, and reason: A unified evaluation of retrieval-augmented generation
4 Pith papers cite this work, alongside 14 external citations. Polarity classification is still indexing.
4
Pith papers citing it
14
external citations · OpenAlex
years
2026 4representative citing papers
LiveBrowseComp shows search agents rely on intrinsic knowledge on standard benchmarks, with scores dropping 25-40 points and closed-book accuracy below 2% on questions about facts from the prior 90 days.
Counterfactual no-search vs forced-search outcomes yield a model-specific oracle that trains search-routing policies, raising macro-F1 from ~0.71 to ~0.82–0.84 on oracle-eligible examples.
CERA fine-tunes a dense retriever with triplet contrastive learning plus attention alignment to human rationales, claiming better retrieval effectiveness and faithfulness on clinical trial reports than Contriever and standard hard-negative baselines.
citing papers explorer
-
LoHoSearch: Benchmarking Long-Horizon Search Agents Beyond the Human Difficulty Ceiling
LoHoSearch is a new benchmark of 544 KG-constructed questions across 11 domains where the strongest search agent scores 34.74% and context strategies add at most 6.8%.
-
LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know?
LiveBrowseComp shows search agents rely on intrinsic knowledge on standard benchmarks, with scores dropping 25-40 points and closed-book accuracy below 2% on questions about facts from the prior 90 days.
-
When Should LLMs Search? Counterfactual Supervision for Search Routing
Counterfactual no-search vs forced-search outcomes yield a model-specific oracle that trains search-routing policies, raising macro-F1 from ~0.71 to ~0.82–0.84 on oracle-eligible examples.
-
Beyond Topical Similarity: Contrastive Evidence Retrieval with Interpretable Attention Alignment in RAG
CERA fine-tunes a dense retriever with triplet contrastive learning plus attention alignment to human rationales, claiming better retrieval effectiveness and faithfulness on clinical trial reports than Contriever and standard hard-negative baselines.