GPT-5 vs Other LLMs in Long Short-Context Performance

Nima Esmi, Maryam Nezhad-Moghaddam, Fatemeh Borhani, Asadollah Shahbahrami, Amin Daemdoost, Georgi Gaydadjiev · Feb 15, 2026 · Citations: 0

Automatic Metrics

Open arXiv Open DOI Find Implementation RSS feed

Abstract

With the significant expansion of the context window in Large Language Models (LLMs), these models are theoretically capable of processing millions of tokens in a single pass. However, research indicates a significant gap between this theoretical capacity and the practical ability of models to robustly utilize information within long contexts, especially in tasks that require a comprehensive understanding of numerous details. This paper evaluates the performance of four state-of-the-art models (Grok-4, GPT-4, Gemini 2.5, and GPT-5) on long short-context tasks. For this purpose, three datasets were used: two supplementary datasets for retrieving culinary recipes and math problems, and a primary dataset of 20K social media posts for depression detection. The results show that as the input volume on the social media dataset exceeds 5K posts (70K tokens), the performance of all models degrades significantly, with accuracy dropping to around 50-53% for 20K posts. Notably, in the GPT-5 model, despite the sharp decline in accuracy, its precision remained high at approximately 95%, a feature that could be highly effective for sensitive applications like depression detection. This research also indicates that the "lost in the middle" problem has been largely resolved in newer models. This study emphasizes the gap between the theoretical capacity and the actual performance of models on complex, high-volume data tasks and highlights the importance of metrics beyond simple accuracy for practical applications.

HFEPX Relevance Assessment

This paper appears adjacent to HFEPX scope (human-feedback/eval), but does not show strong direct protocol evidence in metadata/abstract.

Eval-Fit Score

0/100 • Low

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

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Detected

HFEPX Fit

Adjacent candidate

If you are doing eval pipeline work, start here:

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

Human Data Lens

Uses human feedback: No
Feedback types: None
Rater population: Unknown
Unit of annotation: Unknown
Expertise required: Math
Extraction source: Persisted extraction

Evaluation Lens

Evaluation modes: Automatic Metrics
Agentic eval: None
Quality controls: Not reported
Confidence: 0.35
Flags: low_signal, possible_false_positive

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

accuracyprecision

Research Brief

Deterministic synthesis

The results show that as the input volume on the social media dataset exceeds 5K posts (70K tokens), the performance of all models degrades significantly, with accuracy dropping to around 50-53% for 20K posts. HFEPX signals include Automatic Metrics with confidence 0.35. Updated from current HFEPX corpus.

Generated Mar 5, 2026, 2:21 AM · Grounded in abstract + metadata only

Key Takeaways

The results show that as the input volume on the social media dataset exceeds 5K posts (70K tokens), the performance of all models degrades significantly, with accuracy dropping…
Notably, in the GPT-5 model, despite the sharp decline in accuracy, its precision remained high at approximately 95%, a feature that could be highly effective for sensitive…
Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Researcher Actions

Treat this as method context, then pivot to protocol-specific HFEPX hubs.
Identify benchmark choices from full text before operationalizing conclusions.
Validate metric comparability (accuracy, precision).

Caveats

Generated from title, abstract, and extracted metadata only; full-paper implementation details are not parsed.
Low-signal flag detected: protocol relevance may be indirect.

Recommended Queries

human-eval protocol design pairwise preference data quality inter-rater agreement adjudication

Research Summary

Contribution Summary

The results show that as the input volume on the social media dataset exceeds 5K posts (70K tokens), the performance of all models degrades significantly, with accuracy dropping to around 50-53% for 20K posts.
Notably, in the GPT-5 model, despite the sharp decline in accuracy, its precision remained high at approximately 95%, a feature that could be highly effective for sensitive applications like depression detection.
This study emphasizes the gap between the theoretical capacity and the actual performance of models on complex, high-volume data tasks and highlights the importance of metrics beyond simple accuracy for practical applications.

Why It Matters For Eval

Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

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.
Gap: Benchmark or dataset anchors are present

No benchmark/dataset anchor extracted from abstract.
Pass: Metric reporting is present

Detected: accuracy, precision

Category-Adjacent Papers (Broader Context)

These papers are nearby in arXiv category and useful for broader context, but not necessarily protocol-matched to this paper.

Faster, Cheaper, More Accurate: Specialised Knowledge Tracing Models Outperform LLMs Category Neighbor

Citations: 0 Relevance: 5.35
- Shared arXiv category (cs.CL, cs.AI)
- Shared metric mentions
- Shared terminology (accuracy, llms, performance)
Efficient Self-Evaluation for Diffusion Language Models via Sequence Regeneration Category Neighbor

Citations: 0 Relevance: 4.45
- Shared arXiv category (cs.CL, cs.AI)
- Shared metric mentions
- Shared terminology (accuracy)
Think, But Don't Overthink: Reproducing Recursive Language Models Category Neighbor

Citations: 0 Relevance: 3.75
- Shared arXiv category (cs.CL)
- Shared metric mentions
- Shared terminology (accuracy, llms, performance)
Cross-Family Speculative Prefill: Training-Free Long-Context Compression with Small Draft Models Category Neighbor

Citations: 0 Relevance: 3.30
- Shared arXiv category (cs.CL)
- Shared metric mentions
- Shared terminology (accuracy, performance)
Contextualized Privacy Defense for LLM Agents Category Neighbor

Citations: 0 Relevance: 3.20
- Shared arXiv category (cs.CL, cs.AI)
Guideline-Grounded Evidence Accumulation for High-Stakes Agent Verification Category Neighbor

Citations: 0 Relevance: 3.20
- Shared arXiv category (cs.CL, cs.AI)
ExpGuard: LLM Content Moderation in Specialized Domains Category Neighbor

Citations: 0 Relevance: 2.50
- Shared arXiv category (cs.CL)
- Shared terminology (llms, performance)
MaBERT:A Padding Safe Interleaved Transformer Mamba Hybrid Encoder for Efficient Extended Context Masked Language Modeling Category Neighbor

Citations: 0 Relevance: 2.50
- Shared arXiv category (cs.CL)
- Shared terminology (long, performance)

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote