Assessing LLM Reliability on Temporally Recent Open-Domain Questions

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Krishnappa, Pushwitha, Das, Amit, Jain, Vinija, Mukherjee, Tathagata, Chadha, Aman
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912898756902912
author Krishnappa, Pushwitha
Das, Amit
Jain, Vinija
Mukherjee, Tathagata
Chadha, Aman
author_facet Krishnappa, Pushwitha
Das, Amit
Jain, Vinija
Mukherjee, Tathagata
Chadha, Aman
contents Large Language Models (LLMs) are increasingly deployed for open-domain question answering, yet their alignment with human perspectives on temporally recent information remains underexplored. We introduce RECOM (Reddit Evaluation for Correspondence of Models), a benchmark dataset of 15,000 recent Reddit questions from September 2025 paired with community-derived reference answers. We investigate how four open-source LLMs (Llama3.1-8B, Mistral-7B, Gemma-2-9B, and GPT-OSS-20B) respond to these questions, evaluating alignment using lexical metrics (BLEU, ROUGE), semantic similarity (BERTScore, MoverScore, cosine similarity), and logical inference (NLI). Our central finding is a striking semantic-lexical paradox: all models achieve over 99% cosine similarity with references despite less than 8% BLEU-1 overlap, a 90+ percentage point gap indicating that models preserve meaning through extensive paraphrasing rather than lexical reproduction. MoverScore (51-53%) confirms this pattern, occupying an intermediate position that reflects the optimal transport cost of semantic alignment. Furthermore, model scale does not predict performance: Mistral-7B (7B parameters) outperforms GPT-OSS-20B (20B parameters) across all metrics. NLI analysis reveals that contradiction rates remain below 7%, suggesting models rarely generate content that directly conflicts with human consensus. These findings challenge the reliability of lexical metrics for evaluating abstractive generation and argue for multi-dimensional evaluation frameworks that capture semantic fidelity beyond surface-level text matching. The RECOM dataset is publicly available at https://anonymous.4open.science/r/recom-D4B0
format Preprint
id arxiv_https___arxiv_org_abs_2602_11165
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Assessing LLM Reliability on Temporally Recent Open-Domain Questions
Krishnappa, Pushwitha
Das, Amit
Jain, Vinija
Mukherjee, Tathagata
Chadha, Aman
Computation and Language
Artificial Intelligence
Large Language Models (LLMs) are increasingly deployed for open-domain question answering, yet their alignment with human perspectives on temporally recent information remains underexplored. We introduce RECOM (Reddit Evaluation for Correspondence of Models), a benchmark dataset of 15,000 recent Reddit questions from September 2025 paired with community-derived reference answers. We investigate how four open-source LLMs (Llama3.1-8B, Mistral-7B, Gemma-2-9B, and GPT-OSS-20B) respond to these questions, evaluating alignment using lexical metrics (BLEU, ROUGE), semantic similarity (BERTScore, MoverScore, cosine similarity), and logical inference (NLI). Our central finding is a striking semantic-lexical paradox: all models achieve over 99% cosine similarity with references despite less than 8% BLEU-1 overlap, a 90+ percentage point gap indicating that models preserve meaning through extensive paraphrasing rather than lexical reproduction. MoverScore (51-53%) confirms this pattern, occupying an intermediate position that reflects the optimal transport cost of semantic alignment. Furthermore, model scale does not predict performance: Mistral-7B (7B parameters) outperforms GPT-OSS-20B (20B parameters) across all metrics. NLI analysis reveals that contradiction rates remain below 7%, suggesting models rarely generate content that directly conflicts with human consensus. These findings challenge the reliability of lexical metrics for evaluating abstractive generation and argue for multi-dimensional evaluation frameworks that capture semantic fidelity beyond surface-level text matching. The RECOM dataset is publicly available at https://anonymous.4open.science/r/recom-D4B0
title Assessing LLM Reliability on Temporally Recent Open-Domain Questions
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2602.11165