Skip to content
← Back to explorer

HFEPX Hub

Multilingual Papers (Last 45 Days)

Updated from current HFEPX corpus (Apr 9, 2026). 24 papers are grouped in this hub page.

Read Full Context

Updated from current HFEPX corpus (Apr 9, 2026). 24 papers are grouped in this hub page. Common evaluation modes: Automatic Metrics, Human Eval. Most common rater population: Domain Experts. Common annotation unit: Pairwise. Frequent quality control: Calibration. Frequently cited benchmark: AdvBench. 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 Feb 27, 2026.

Papers: 24 Last published: Feb 27, 2026 Global RSS Tag RSS
MultilingualLast 45d

Researcher Quick Triage

This hub is best used for protocol triage and replication planning from abstract-level evidence. Quality band: Developing .

All Sampled Papers (24) Replication-Ready Only (2)

High-Signal Coverage

100.0%

24 / 24 sampled papers are not low-signal flagged.

Replication-Ready Set

2

Benchmark + metric + eval mode explicitly present.

Judge/Human Comparability

0

Papers containing both `human_eval` and `llm_as_judge`.

  • 2 papers are replication-ready (benchmark + metric + explicit evaluation mode).
  • 0 papers support judge-vs-human agreement analysis.
  • 2 papers report explicit quality controls (calibration/adjudication/IAA).

Primary action: Use this page for scouting only; collect additional papers before attempting replication-critical comparisons.

Need evaluators for this research workflow?

Post a Job →

Why This Matters For Eval Research

  • 62.5% of papers report explicit human-feedback signals, led by pairwise preferences.
  • automatic metrics appears in 58.3% of papers in this hub.
  • AdvBench is a recurring benchmark anchor for cross-paper comparisons in this page.

Protocol Takeaways

  • Most common quality-control signal is rater calibration (8.3% of papers).
  • Rater context is mostly domain experts, and annotation is commonly pairwise annotation; use this to scope replication staffing.
  • Compare papers that report both human_eval and llm_as_judge to quantify judge-human agreement drift.

Benchmark Interpretation

  • AdvBench appears in 4.2% of hub papers (1/24); use this cohort for benchmark-matched comparisons.
  • Jbf-Eval appears in 4.2% of hub papers (1/24); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • accuracy is reported in 33.3% of hub papers (8/24); compare with a secondary metric before ranking methods.
  • bleu is reported in 12.5% of hub papers (3/24); compare with a secondary metric before ranking methods.
Researcher Checklist (Expanded)

Researcher Checklist

  • Strong: Papers with explicit human feedback

    Coverage is strong (62.5% vs 45% target).

  • Gap: Papers reporting quality controls

    Coverage is a replication risk (8.3% vs 30% target).

  • Gap: Papers naming benchmarks/datasets

    Coverage is a replication risk (8.3% vs 35% target).

  • Strong: Papers naming evaluation metrics

    Coverage is strong (62.5% vs 35% target).

  • Moderate: Papers with known rater population

    Coverage is usable but incomplete (25% vs 35% target).

  • Gap: Papers with known annotation unit

    Coverage is a replication risk (20.8% vs 35% target).

Strengths

  • Strong human-feedback signal (62.5% of papers).
  • Contains both human-eval and LLM-as-judge protocols for head-to-head methodology comparison.
  • Agentic evaluation appears in 33.3% of papers.

Known Gaps

  • Only 8.3% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Annotation unit is under-specified (20.8% coverage).
  • Benchmark coverage is thin (8.3% of papers mention benchmarks/datasets).

Suggested Next Analyses

  • Compare papers that report both human_eval and llm_as_judge to quantify judge-human agreement drift.
  • Stratify by benchmark (AdvBench vs Jbf-Eval) before comparing methods.
  • Track metric sensitivity by reporting both accuracy and bleu.
  • Add inter-annotator agreement checks when reproducing these protocols.
Recommended Queries (Expanded)

Recommended Queries

Judge vs Human Agreement Benchmark Slice: AdvBench Metric Slice: accuracy Recent High-Signal Papers

Start with These 3

Use these when you need one protocol anchor, one benchmark anchor, and one recent comparison point before reading the wider hub.

Start Here (Best First 6)

Ranked for protocol completeness (human signal, benchmark + metric anchors, quality controls, and judge/human overlap).

Protocol Matrix (Top 12)

Use this to quickly compare protocol ingredients instead of scanning long prose.

Paper HF Signal Eval Modes Benchmarks Metrics QC
Jailbreak Foundry: From Papers to Runnable Attacks for Reproducible Benchmarking

Feb 27, 2026

 Yes
Red Team
Llm As Judge AdvBench , Jbf Eval Success rate , Jailbreak success rate Not Reported
A Multi-Stage Validation Framework for Trustworthy Large-scale Clinical Information Extraction using Large Language Models

Apr 7, 2026

 Yes
Expert Verification
Automatic Metrics Not Reported F1 , Agreement Calibration , Adjudication
Blinded Radiologist and LLM-Based Evaluation of LLM-Generated Japanese Translations of Chest CT Reports: Comparative Study

Apr 2, 2026

 Yes
Pairwise Preference
Llm As Judge , Automatic Metrics Not Reported Accuracy Not Reported
Semantic Alignment across Ancient Egyptian Language Stages via Normalization-Aware Multitask Learning

Mar 25, 2026

 Yes
Pairwise Preference
Automatic Metrics Not Reported Accuracy Not Reported
Do Compact SSL Backbones Matter for Audio Deepfake Detection? A Controlled Study with RAPTOR

Mar 6, 2026

 Yes
Pairwise Preference
Not Reported Not Reported Not Reported Calibration
LIT-RAGBench: Benchmarking Generator Capabilities of Large Language Models in Retrieval-Augmented Generation

Mar 6, 2026

 No
Not Reported
Llm As Judge , Automatic Metrics Lit Ragbench Accuracy Not Reported
A prospective clinical feasibility study of a conversational diagnostic AI in an ambulatory primary care clinic

Mar 9, 2026

 Yes
Expert Verification
Automatic Metrics Not Reported Accuracy , Cost Not Reported
MEDSYN: Benchmarking Multi-EviDence SYNthesis in Complex Clinical Cases for Multimodal Large Language Models

Feb 25, 2026

 Yes
Expert Verification
Automatic Metrics Not Reported Accuracy Not Reported
Obscure but Effective: Classical Chinese Jailbreak Prompt Optimization via Bio-Inspired Search

Feb 26, 2026

 Yes
Red Team
Automatic Metrics Not Reported Accuracy , Conciseness Not Reported
To Write or to Automate Linguistic Prompts, That Is the Question

Mar 26, 2026

 Yes
Expert Verification
Not Reported Not Reported Not Reported Not Reported
Plausibility as Commonsense Reasoning: Humans Succeed, Large Language Models Do not

Apr 6, 2026

 Yes
Pairwise Preference
Not Reported Not Reported Not Reported Not Reported
Cross-Preference Learning for Sentence-Level and Context-Aware Machine Translation

Mar 26, 2026

 Yes
Pairwise Preference
Not Reported Not Reported Not Reported Not Reported

Protocol Diff (Top Papers)

Fast side-by-side comparison for the highest-ranked papers in this hub.

Signal Jailbreak Foundry: From Papers to Runnable Attacks… A Multi-Stage Validation Framework for Trustworthy… Blinded Radiologist and LLM-Based Evaluation of LLM…
Human Feedback Red TeamExpert VerificationPairwise Preference
Evaluation Modes Llm As JudgeAutomatic MetricsLlm As Judge, Automatic Metrics
Benchmarks AdvBench, Jbf EvalNot reportedNot reported
Metrics Success rate, Jailbreak success rateF1, AgreementAccuracy
Quality Controls Not reportedCalibration, AdjudicationNot reported
Rater Population UnknownDomain ExpertsDomain Experts
Annotation Unit UnknownUnknownPairwise
Suggested Reading Order (Extended)

This section is intentionally expanded only when needed; use “Start Here” above for a faster pass.

Suggested Reading Order

  1. A Multi-Stage Validation Framework for Trustworthy Large-scale Clinical Information Extraction using Large Language Models

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics + expert verification. Focus: f1. Abstract: Conventional evaluation methods rely heavily on annotation-intensive reference standards or.

  2. Plausibility as Commonsense Reasoning: Humans Succeed, Large Language Models Do not

    Start here for detailed protocol reporting and quality-control evidence. Signals: pairwise preferences. Abstract: We then evaluate Turkish and multilingual LLMs in a parallel preference-based setup that compares matched.

  3. Blinded Radiologist and LLM-Based Evaluation of LLM-Generated Japanese Translations of Chest CT Reports: Comparative Study

    Start here for detailed protocol reporting and quality-control evidence. Signals: LLM-as-judge + pairwise preferences. Focus: accuracy. Abstract: Objective: To evaluate the educational suitability of LLM-generated Japanese translations of.

  4. Terminology Rarity Predicts Catastrophic Failure in LLM Translation of Low-Resource Ancient Languages: Evidence from Ancient Greek

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: human evaluation. Focus: bleu. Abstract: This study presents the first systematic, reference-free human evaluation of large language.

  5. Voxtral TTS

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: human evaluation. Focus: win rate. Abstract: In human evaluations conducted by native speakers, Voxtral TTS is preferred.

  6. Jailbreak Foundry: From Papers to Runnable Attacks for Reproducible Benchmarking

    Include an LLM-as-judge paper to test judge design and agreement assumptions. Signals: LLM-as-judge + red-team protocols. Focus: AdvBench / success rate. Abstract: This system enables a standardized AdvBench.

  7. Do Compact SSL Backbones Matter for Audio Deepfake Detection? A Controlled Study with RAPTOR

    Adds evaluation protocol evidence with pairwise preferences for broader protocol coverage within this hub. Signals: pairwise preferences. Abstract: We present RAPTOR, Representation Aware Pairwise-gated Transformer for Out-of-domain Recognition.

  8. LIT-RAGBench: Benchmarking Generator Capabilities of Large Language Models in Retrieval-Augmented Generation

    Adds LLM-as-judge for broader protocol coverage within this hub. Signals: LLM-as-judge. Focus: lit-ragbench / accuracy. Abstract: We use LLM-as-a-Judge for scoring and report category-wise and overall accuracy.

Known Limitations

Known Limitations

  • Only 8.3% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Annotation unit is under-specified (20.8% coverage).
  • Narrative synthesis is grounded in metadata and abstracts only; full-paper implementation details are not parsed.
Research Utility Snapshot (Detailed)

Research Utility Snapshot

Human Feedback Mix

  • Pairwise Preference (9)
  • Expert Verification (4)
  • Red Team (2)

Evaluation Modes

  • Automatic Metrics (14)
  • Human Eval (3)
  • Llm As Judge (3)
  • Simulation Env (1)

Top Benchmarks

  • AdvBench (1)
  • Jbf Eval (1)
  • Lit Ragbench (1)

Top Metrics

  • Accuracy (8)
  • Bleu (3)
  • F1 (3)
  • Agreement (2)

Rater Population Mix

  • Domain Experts (6)

Quality Controls

  • Calibration (2)
  • Adjudication (1)
Coverage diagnostics (sample-based): human-feedback 62.5% · benchmarks 8.3% · metrics 62.5% · quality controls 8.3%.

Top Papers

Related Hubs

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

Post a Job Get a Quote