---
source_url: "https://machinelearning.apple.com/research/search-augmented?utm_source=openai"
title: Over-Searching in Search-Augmented Large Language Models - Apple Machine Learning Research
mirrored_at: 2026-08-14T03:33:32.639Z
host: machinelearning.apple.com
cited_in_42a: true
mirror_canonical: "https://index.42a.ai/machinelearning.apple.com/research/search-augmented__q__utm_source_openai"
---

> **Original source:** https://machinelearning.apple.com/research/search-augmented?utm_source=openai

AuthorsRoy Xie†\*\*, Deepak Gopinath\*\*, David Qiu\*\*, Dong Lin, Haitian Sun, Saloni Potdar, Bhuwan Dhingra†

Search-augmented large language models (LLMs) excel at knowledge-intensive tasks by integrating external retrieval. However, they often over-search – unnecessarily invoking search tool even when it does not improve response quality, which leads to computational inefficiency and hallucinations by incorporating irrelevant context. In this work, we conduct a systematic evaluation of over-searching across multiple dimensions, including query types, model categories, retrieval conditions, and multi-turn conversations. Our finding shows: (i) search generally improves answer accuracy on answerable queries but harms abstention on unanswerable ones; (ii) over-searching is more pronounced in complex reasoning models and deep research systems, is exacerbated by noisy retrieval, and compounds across turns in multi-turn conversations; and (iii) the composition of retrieved evidence is crucial, as the presence of negative evidence improves abstention. To quantify over-searching, we introduce Tokens Per Correctness (TPC), an evaluation metric that captures the performance-cost trade-off for search-augmented LLMs. Lastly, we investigate mitigation approaches at both the query and retrieval levels and release the OverSearchQA benchmark to foster continued research into efficient search-augmented LLMs.

-   † Duke University
-   \*\* Work done while at Apple

## Related readings and updates.

Multimodal Large Language Models (MLLMs) in real-world applications require access to external knowledge sources and must remain responsive to the dynamic and ever-changing real-world information in order to address information-seeking and knowledge-intensive user queries. Existing approaches, such as retrieval augmented generation (RAG) methods, search agents, and search equipped MLLMs, often suffer from rigid pipelines, excessive search calls,…

[Read more](https://machinelearning.apple.com/research/r1)

This paper was accepted at the UncertaiNLP workshop at EACL 2024.

Large language models (LLMs) have the remarkable ability to solve new tasks with just a few examples, but they need access to the right tools. Retrieval Augmented Generation (RAG) addresses this problem by retrieving a list of relevant tools for a given task. However, RAG’s tool retrieval step requires all the required information to be explicitly present in the query. This is a…

[Read more](https://machinelearning.apple.com/research/context-tuning-retrieval)