Not All Needles Are Found: How Fact Distribution and Don't Make It Up Prompts Shape Literal Extraction, Logical Inference, and Hallucination Risks in Long-Context LLMs

TL;DR

This study investigates how fact placement, distribution, and prompt design affect the performance, inference, and hallucination risks of long-context large language models across various tasks and models.

Contribution

It introduces a comprehensive benchmark analyzing literal extraction, inference, and hallucination, revealing how context length and fact distribution impact model reliability.

Findings

Longer contexts do not always improve performance and can harm accuracy when evidence is dispersed.

Model robustness varies significantly, with some models degrading under realistic long-context conditions.

Anti-hallucination prompts can reduce hallucinations but may also decrease extraction and inference accuracy.

Abstract

Large language models (LLMs) increasingly support very long input contexts. Yet it remains unclear how reliably they extract and infer information at scale. Performance varies with context length and strongly interacts with how information is distributed in real-world corpora. Motivated by these observations, we study how fact placement, corpus-level fact distributions, and Don't Make It Up prompts influence model behavior. We introduce an extended needle-in-a-haystack benchmark across four production-scale models: Gemini-2.5-flash, ChatGPT-5-mini, Claude-4.5-haiku, and Deepseek-v3.2-chat. Unlike prior work, we separately evaluate literal extraction, logical inference, and hallucination risk. Our study considers both positional effects and realistic distributions of evidence across long contexts, as well as prompts that explicitly discourage fabrication. We find that longer contexts alone do not guarantee better performance and can be detrimental when relevant evidence is diluted or widely dispersed. Performance varies substantially across models: some show severe degradation under realistic conditions, while others remain more robust at longer context lengths. Anti-hallucination (AH) instructions can make some models overly conservative, sharply reducing accuracy in literal extraction and logical inference. While we do not directly compare retrieval-augmented generation (RAG) and cache-augmented generation (CAG), our results suggest many failures stem from ineffective context utilization. Models often struggle to identify and prioritize relevant information even when it is present. These findings have direct practical implications, as enterprise workflows increasingly involve pasting large volumes of unfiltered documents into LLM prompts. Effective context length and model-specific robustness to long contexts are therefore critical for reliable LLM deployment in research and business.