Skip to content
OpenTrain AIFor AI Companies
← Back to explorer

How Far Can You Get Without a GPU? A Systematic Benchmark of Lightweight Hallucination Detection Across Question Answering, Dialogue, and Summarisation

Kriti Faujdar, Smit Kadvani · Jun 29, 2026 · Citations: 0

Automatic Metrics General Tool Use
Open arXiv Find Implementation RSS feed Shortlist (0)

How to use this page

Moderate trust

Use this for comparison and orientation, not as your only source.

Best use

Background context only

What to verify

Validate the evaluation procedure and quality controls in the full paper before operational use.

Evidence quality

Moderate

Derived from extracted protocol signals and abstract evidence.

Abstract

Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale. The most accurate detection methods depend on GPU-intensive inference, proprietary API calls, or white-box access to the generating model. This puts them out of reach for resource-constrained researchers and practitioners. In this paper, we explore a practical alternative: how well can hallucination detection perform using only lightweight, CPU-feasible methods built on publicly available models? We systematically benchmark five such methods: ROUGE-L, semantic similarity, BERTScore, a Natural Language Inference (NLI) detector based on a FEVER-trained DeBERTa model, and a score-level ensemble of similarity and NLI. We evaluate them across all three tasks of the HaluEval benchmark: question answering (QA), dialogue, and summarisation. We calibrate each method on a held-out validation split and evaluate it on 2,000 test instances per task. We find that no single method dominates and performance is highly task-dependent. The ensemble performs best on QA (F1 = 0.792, AUC-ROC = 0.873), the NLI detector leads on dialogue (AUC-ROC = 0.713), and all five methods degrade to near-random performance on summarisation (AUC-ROC between 0.469 and 0.574). This task-dependence and the systematic failure on summarisation map the practical frontier of GPU-free hallucination detection. They give practical guidance for method selection under computational constraints. All experiments run on a standard laptop CPU using public models.

Should You Rely On This Paper?

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Best use

Background context only

Use if you need

A benchmark-and-metrics comparison anchor.

Main weakness

No major weakness surfaced.

Trust level

Moderate

Usefulness score

25/100 • Low

Treat as adjacent context, not a core eval-method reference.

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Detected

Usefulness for eval research

Adjacent candidate

Extraction confidence 55%

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

What We Could Verify

These are the protocol signals we could actually recover from the available paper metadata. Use them to decide whether this paper is worth deeper reading.

Human Feedback Types

missing

None explicit

No explicit feedback protocol extracted.

"Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale."

Evaluation Modes

strong

Automatic Metrics

Includes extracted eval setup.

"Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale."

Quality Controls

missing

Not reported

No explicit QC controls found.

"Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale."

Benchmarks / Datasets

strong

FEVER, Halueval

Useful for quick benchmark comparison.

"We systematically benchmark five such methods: ROUGE-L, semantic similarity, BERTScore, a Natural Language Inference (NLI) detector based on a FEVER-trained DeBERTa model, and a score-level ensemble of similarity and NLI."

Reported Metrics

strong

F1, Rouge, Auroc, Bertscore

Useful for evaluation criteria comparison.

"We systematically benchmark five such methods: ROUGE-L, semantic similarity, BERTScore, a Natural Language Inference (NLI) detector based on a FEVER-trained DeBERTa model, and a score-level ensemble of similarity and NLI."

Human Feedback Details

  • Uses human feedback: No
  • Feedback types: None
  • Rater population: Not reported
  • Expertise required: General

Evaluation Details

  • Evaluation modes: Automatic Metrics
  • Agentic eval: Tool Use
  • Quality controls: Not reported
  • Evidence quality: Moderate
  • Use this page as: Background context only

Protocol And Measurement Signals

Benchmarks / Datasets

FEVERHalueval

Reported Metrics

f1rougeaurocbertscore

Research Brief

Metadata summary

Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale.

Based on abstract + metadata only. Check the source paper before making high-confidence protocol decisions.

Key Takeaways

  • Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale.
  • The most accurate detection methods depend on GPU-intensive inference, proprietary API calls, or white-box access to the generating model.
  • This puts them out of reach for resource-constrained researchers and practitioners.

Researcher Actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Validate inferred eval signals (Automatic metrics, Tool-use evaluation) against the full paper.
  • Use related-paper links to find stronger protocol-specific references.

Caveats

  • Generated from abstract + metadata only; no PDF parsing.
  • Signals below are heuristic and may miss details reported outside the abstract.

Recommended Queries

F1

Research Summary

Contribution Summary

  • We systematically benchmark five such methods: ROUGE-L, semantic similarity, BERTScore, a Natural Language Inference (NLI) detector based on a FEVER-trained DeBERTa model, and a score-level ensemble of similarity and NLI.
  • We evaluate them across all three tasks of the HaluEval benchmark: question answering (QA), dialogue, and summarisation.
  • The ensemble performs best on QA (F1 = 0.792, AUC-ROC = 0.873), the NLI detector leads on dialogue (AUC-ROC = 0.713), and all five methods degrade to near-random performance on summarisation (AUC-ROC between 0.469 and 0.574).

Why It Matters For Eval

  • We systematically benchmark five such methods: ROUGE-L, semantic similarity, BERTScore, a Natural Language Inference (NLI) detector based on a FEVER-trained DeBERTa model, and a score-level ensemble of similarity and NLI.
  • We evaluate them across all three tasks of the HaluEval benchmark: question answering (QA), dialogue, and summarisation.

Researcher Checklist

  • Gap: Human feedback protocol is explicit

    No explicit human feedback protocol detected.

  • Pass: Evaluation mode is explicit

    Detected: Automatic Metrics

  • Gap: Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Pass: Benchmark or dataset anchors are present

    Detected: FEVER, Halueval

  • Pass: Metric reporting is present

    Detected: f1, rouge, auroc, bertscore

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.