Skip to content
OpenTrain AIFor AI Companies
← Back to explorer

HFEPX Benchmark Hub

BFCL or GPQA Benchmark Papers

Updated from current HFEPX corpus (Sep 1, 2026). 43 papers are grouped in this benchmark page.

Read Full Context

Updated from current HFEPX corpus (Sep 1, 2026). 43 papers are grouped in this benchmark page. Common evaluation modes: Automatic Metrics, Llm As Judge. Common annotation unit: Trajectory. Frequent quality control: Adjudication. Frequently cited benchmark: GPQA. Common metric signal: accuracy. Use this page to compare protocol setup, judge behavior, and labeling design decisions before running new eval experiments. Newest paper in this set is from May 30, 2026.

Papers: 43 Last published: May 30, 2026 Global RSS

Researcher Quick Triage

Use this page for benchmark-matched method comparisons and eval protocol selection. Quality band: High .

High-Signal Coverage

100.0%

43 / 43 sampled papers are not low-signal flagged.

Replication-Ready Set

14

Papers with explicit benchmark + metric + eval mode fields.

Quality Controls

2.3%

1 papers report calibration/adjudication/IAA controls.

  • 15 papers explicitly name benchmark datasets in the sampled set.
  • 14 papers report at least one metric term in metadata extraction.
  • Start with the ranked shortlist below before reading all papers.

Primary action: Start with the top 2 benchmark-matched papers, then compare evaluation modes in the protocol matrix.

Why This Matters (Expanded)

Why This Matters For Eval Research

  • 11.6% of papers report explicit human-feedback signals, led by pairwise preferences.
  • automatic metrics appears in 30.2% of papers in this hub.
  • GPQA is a recurring benchmark anchor for cross-paper comparisons in this page.
Protocol Notes (Expanded)

Protocol Takeaways

  • Most common quality-control signal is adjudication (2.3% of papers).
  • Rater context is mostly unspecified rater pools, and annotation is commonly trajectory-level annotation; use this to scope replication staffing.
  • Pair this hub with a human_eval-heavy hub to validate judge-model calibration.

Benchmark Interpretation

  • GPQA appears in 62.8% of hub papers (27/43); use this cohort for benchmark-matched comparisons.
  • BFCL appears in 37.2% of hub papers (16/43); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • accuracy is reported in 39.5% of hub papers (17/43); compare with a secondary metric before ranking methods.
  • cost is reported in 18.6% of hub papers (8/43); compare with a secondary metric before ranking methods.

Start Here (Benchmark-Matched First 6)

Ranked by protocol completeness so you can quickly find papers suitable for comparison studies.

Protocol Matrix (Top 10)

Compare protocol ingredients quickly before deep-reading full papers.

Paper Eval Modes Human Feedback Metrics Quality Controls
When Self-Consistency Backfires: Majority Vote Hurts the Majority of Hard Science Problems for Small LLMs

Aug 11, 2026

 Automatic Metrics Demonstrations Accuracy Adjudication
Decomposing Wrong-Consensus Agreement in LLM Self-Consistency

Aug 19, 2026

 Automatic Metrics Pairwise Preference Accuracy Not reported
Skill or Skip? Learning Selective Skill Invocation in Agentic Tasks via Dual-Granularity Preference Learning

May 30, 2026

 Simulation Env Pairwise Preference Precision, Task success Not reported
Breaking MCP with Function Hijacking Attacks: Novel Threats for Function Calling and Agentic Models

Apr 22, 2026

 Automatic Metrics Pairwise Preference, Red Team Jailbreak success rate Not reported
Beyond Single-Turn Confidence: Trajectory-Adapted Uncertainty Quantification for LLM Agents

Aug 12, 2026

 Automatic Metrics Not reported Cost Not reported
Brief Is Better: Non-Monotonic Chain-of-Thought Budget Effects in Function-Calling Language Agents

Apr 2, 2026

 Automatic Metrics Not reported Accuracy Not reported
SkillX: Automatically Constructing Skill Knowledge Bases for Agents

Apr 6, 2026

 Automatic Metrics Not reported Task success Not reported
Top-b: Entropic Regulation of Relative Probability Bands in Autoregressive Language Processes

Mar 15, 2026

 Automatic Metrics Not reported Accuracy Not reported
D-COT: Disciplined Chain-of-Thought Learning for Efficient Reasoning in Small Language Models

Feb 25, 2026

 Automatic Metrics Not reported Accuracy Not reported
DeepPrune: Parallel Scaling without Inter-trace Redundancy

Oct 9, 2025

 Llm As Judge, Automatic Metrics Not reported Accuracy, Auroc Not reported
Researcher Workflow (Detailed)

Checklist

  • Gap: Papers with explicit human feedback

    Coverage is a replication risk (11.6% vs 45% target).

  • Gap: Papers reporting quality controls

    Coverage is a replication risk (2.3% vs 30% target).

  • Strong: Papers naming benchmarks/datasets

    Coverage is strong (100% vs 35% target).

  • Strong: Papers naming evaluation metrics

    Coverage is strong (72.1% vs 35% target).

  • Gap: Papers with known rater population

    Coverage is a replication risk (0% vs 35% target).

  • Gap: Papers with known annotation unit

    Coverage is a replication risk (14% vs 35% target).

Strengths

  • Most papers provide measurable evaluation context (100% benchmarks, 72.1% metrics).
  • Agentic evaluation appears in 25.6% of papers.

Known Gaps

  • Only 2.3% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (0% coverage).
  • Annotation unit is under-specified (14% coverage).

Suggested Next Analyses

  • Pair this hub with a human_eval-heavy hub to validate judge-model calibration.
  • Stratify by benchmark (GPQA vs BFCL) before comparing methods.
  • Track metric sensitivity by reporting both accuracy and cost.
  • Add inter-annotator agreement checks when reproducing these protocols.

Recommended Queries

LLM-as-Judge ProtocolsBenchmark Slice: GPQAMetric Slice: accuracyRecent High-Signal Papers
Known Limitations
  • Only 2.3% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (0% coverage).
  • Narrative synthesis is grounded in metadata and abstracts only; full-paper implementation details are not parsed.
Research Utility Snapshot (Detailed)

Evaluation Modes

  • Automatic Metrics (13)
  • Llm As Judge (1)
  • Simulation Env (1)

Human Feedback Mix

  • Pairwise Preference (3)
  • Demonstrations (2)
  • Red Team (1)

Top Benchmarks

  • GPQA (27)
  • BFCL (16)
  • AIME (13)
  • MMLU (7)

Top Metrics

  • Accuracy (17)
  • Cost (8)
  • Task success (3)
  • Agreement (2)

Top Papers On This Benchmark

Related Benchmark Hubs

BFCL or HumanEval+ Benchmark Papers (21) BFCL or MATH-500 Benchmark Papers (25) GPQA or HumanEval+ Benchmark Papers (22) GPQA or MATH-500 Benchmark Papers (25) HumanEval+ or MATH-500 Benchmark Papers (24) MATH-500 Benchmark Papers (15) AIME or AlpacaEval or MMLU Benchmark Papers (60) AIME or BFCL or MMLU Benchmark Papers (59) AIME or GSM8K or MMLU Benchmark Papers (72) AIME or HotpotQA or MMLU Benchmark Papers (62) AIME or HumanEval+ or MMLU Benchmark Papers (57)