Skip to content
← Back to explorer

HFEPX Hub

Pairwise Preference Papers (Last 45 Days)

Updated from current HFEPX corpus (Apr 17, 2026). 111 papers are grouped in this hub page.

Read Full Context

Updated from current HFEPX corpus (Apr 17, 2026). 111 papers are grouped in this hub page. Common evaluation modes: Automatic Metrics, Llm As Judge. Most common rater population: Domain Experts. Common annotation unit: Pairwise. Frequent quality control: Calibration. Frequently cited benchmark: AlpacaEval. 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 Apr 8, 2026.

Papers: 111 Last published: Apr 8, 2026 Global RSS Tag RSS
Pairwise PreferenceLast 45d

Researcher Quick Triage

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

Analysis blocks below are computed from the currently loaded sample (60 of 111 total papers in this hub).

High-Signal Coverage

100.0%

60 / 60 sampled papers are not low-signal flagged.

Replication-Ready Set

9

Benchmark + metric + eval mode explicitly present.

Judge/Human Comparability

0

Papers containing both `human_eval` and `llm_as_judge`.

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

Primary action: Start with the top 2 papers in “Start Here”, then validate assumptions in the protocol matrix.

Need evaluators for this research workflow?

Post a Job →

Why This Matters For Eval Research

  • 100% of papers report explicit human-feedback signals, led by pairwise preferences.
  • automatic metrics appears in 36.9% of papers in this hub.
  • AlpacaEval is a recurring benchmark anchor for cross-paper comparisons in this page.

Protocol Takeaways

  • Most common quality-control signal is rater calibration (1.8% 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

  • AlpacaEval appears in 1.8% of hub papers (2/111); use this cohort for benchmark-matched comparisons.
  • IFEval appears in 1.8% of hub papers (2/111); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • accuracy is reported in 20.7% of hub papers (23/111); compare with a secondary metric before ranking methods.
  • cost is reported in 8.1% of hub papers (9/111); compare with a secondary metric before ranking methods.
Researcher Checklist (Expanded)

Researcher Checklist

  • Strong: Papers with explicit human feedback

    Coverage is strong (100% vs 45% target).

  • Gap: Papers reporting quality controls

    Coverage is a replication risk (3.6% vs 30% target).

  • Gap: Papers naming benchmarks/datasets

    Coverage is a replication risk (16.2% vs 35% target).

  • Strong: Papers naming evaluation metrics

    Coverage is strong (42.3% vs 35% target).

  • Gap: Papers with known rater population

    Coverage is a replication risk (9.9% vs 35% target).

  • Strong: Papers with known annotation unit

    Coverage is strong (41.4% vs 35% target).

Strengths

  • Strong human-feedback signal (100% of papers).
  • Contains both human-eval and LLM-as-judge protocols for head-to-head methodology comparison.

Known Gaps

  • Only 3.6% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (9.9% coverage).
  • Benchmark coverage is thin (16.2% 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 (AlpacaEval vs IFEval) before comparing methods.
  • Track metric sensitivity by reporting both accuracy and cost.
Recommended Queries (Expanded)

Recommended Queries

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
Personalized RewardBench: Evaluating Reward Models with Human Aligned Personalization

Apr 8, 2026

Yes Human Eval , Automatic Metrics Rewardbench Accuracy , Helpfulness Not Reported
Do Phone-Use Agents Respect Your Privacy?

Apr 1, 2026

Yes Automatic Metrics APPS , Myphonebench Task success Not Reported
ClimateCheck 2026: Scientific Fact-Checking and Disinformation Narrative Classification of Climate-related Claims

Mar 27, 2026

Yes Automatic Metrics Codabench Recall , Recall@k Not Reported
Stabilizing Iterative Self-Training with Verified Reasoning via Symbolic Recursive Self-Alignment

Mar 23, 2026

Yes Automatic Metrics GSM8K Accuracy Not Reported
DSPA: Dynamic SAE Steering for Data-Efficient Preference Alignment

Mar 23, 2026

Yes Automatic Metrics MT Bench , AlpacaEval Accuracy Not Reported
CausalRM: Causal-Theoretic Reward Modeling for RLHF from Observational User Feedbacks

Mar 19, 2026

Yes Automatic Metrics Harmbench Not Reported Not Reported
From Consensus to Split Decisions: ABC-Stratified Sentiment in Holocaust Oral Histories

Mar 30, 2026

Yes Automatic Metrics Not Reported Kappa , Agreement Inter Annotator Agreement Reported
Measuring Faithfulness Depends on How You Measure: Classifier Sensitivity in LLM Chain-of-Thought Evaluation

Mar 20, 2026

Yes Automatic Metrics Not Reported Kappa , Faithfulness Inter Annotator Agreement Reported
Modeling and Benchmarking Spoken Dialogue Rewards with Modality and Colloquialness

Mar 16, 2026

Yes Automatic Metrics Esdr Bench Accuracy Not Reported
$V_1$: Unifying Generation and Self-Verification for Parallel Reasoners

Mar 4, 2026

Yes Automatic Metrics SWE Bench , AIME Pass@1 Not Reported
AILS-NTUA at SemEval-2026 Task 12: Graph-Based Retrieval and Reflective Prompting for Abductive Event Reasoning

Mar 4, 2026

Yes Automatic Metrics Semeval Accuracy Not Reported
VehicleMemBench: An Executable Benchmark for Multi-User Long-Term Memory in In-Vehicle Agents

Mar 25, 2026

Yes Simulation Env Vehiclemembench Not Reported Not Reported

Protocol Diff (Top Papers)

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

Signal Personalized RewardBench: Evaluating Reward Models… Do Phone-Use Agents Respect Your Privacy? ClimateCheck 2026: Scientific Fact-Checking and Dis…
Human Feedback Pairwise Preference, Rubric RatingPairwise PreferencePairwise Preference
Evaluation Modes Human Eval, Automatic MetricsAutomatic MetricsAutomatic Metrics
Benchmarks RewardbenchAPPS, MyphonebenchCodabench
Metrics Accuracy, HelpfulnessTask successRecall, Recall@k
Quality Controls Not reportedNot reportedNot reported
Rater Population UnknownUnknownUnknown
Annotation Unit PairwiseUnknownUnknown
Suggested Reading Order (Extended)

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

Suggested Reading Order

  1. 152ca5cb-d3d6-4ca1-b4f6-23136013d236

    Start here for detailed protocol reporting and quality-control evidence.

  2. HyperMem: Hypergraph Memory for Long-Term Conversations

    Start here for detailed protocol reporting and quality-control evidence. Signals: LLM-as-judge + pairwise preferences. Focus: accuracy. Abstract: However, existing approaches as Retrieval-Augmented Generation (RAG) and graph-based memory mostly.

  3. Self-Debias: Self-correcting for Debiasing Large Language Models

    Start here for detailed protocol reporting and quality-control evidence. Signals: pairwise preferences. Abstract: Unlike standard preference optimization which applies broad penalties, Self-Debias employs a fine-grained trajectory-level objective subject.

  4. Personalized RewardBench: Evaluating Reward Models with Human Aligned Personalization

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: human evaluation + pairwise preferences. Focus: Rewardbench / accuracy. Abstract: While benchmarks for general response quality are.

  5. PONTE: Personalized Orchestration for Natural Language Trustworthy Explanations

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: human evaluation + pairwise preferences. Focus: agreement. Abstract: It combines: (i) a low-dimensional preference model capturing stylistic.

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

    Include an LLM-as-judge paper to test judge design and agreement assumptions. Signals: LLM-as-judge + pairwise preferences. Focus: accuracy. Abstract: Objective: To evaluate the educational suitability of LLM-generated Japanese.

  7. Build, Judge, Optimize: A Blueprint for Continuous Improvement of Multi-Agent Consumer Assistants

    Include an LLM-as-judge paper to test judge design and agreement assumptions. Signals: LLM-as-judge + pairwise preferences. Abstract: Grocery shopping further amplifies these difficulties, as user requests are often.

  8. VehicleMemBench: An Executable Benchmark for Multi-User Long-Term Memory in In-Vehicle Agents

    Adds simulation environments with pairwise preferences for broader protocol coverage within this hub. Signals: simulation environments + pairwise preferences. Focus: Vehiclemembench. Abstract: This evolution requires agents to continuously.

Known Limitations

Known Limitations

  • Only 3.6% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (9.9% 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 (111)
  • Rubric Rating (6)
  • Expert Verification (3)
  • Critique Edit (1)

Evaluation Modes

  • Automatic Metrics (41)
  • Llm As Judge (11)
  • Simulation Env (4)
  • Human Eval (3)

Top Benchmarks

  • AlpacaEval (2)
  • IFEval (2)
  • AIME (1)
  • APPS (1)

Top Metrics

  • Accuracy (23)
  • Cost (9)
  • Relevance (8)
  • Coherence (7)

Rater Population Mix

  • Domain Experts (10)
  • Mixed (1)

Quality Controls

  • Calibration (2)
  • Inter Annotator Agreement Reported (2)
Coverage diagnostics (sample-based): human-feedback 100.0% · benchmarks 25.0% · metrics 65.0% · quality controls 5.0%.

Top Papers

Related Hubs

Get Started

Join the #1 Platform for AI Training Talent

Where top AI builders and expert AI Trainers connect to build the future of AI.
Self-Service
Post a Job
Post your project and get a shortlist of qualified AI Trainers and Data Labelers. Hire and manage your team in the tools you already use.
Managed Service
For Large Projects
Done-for-You
We recruit, onboard, and manage a dedicated team inside your tools. End-to-end operations for large or complex projects.
For Freelancers
Join as an AI Trainer
Find AI training and data labeling projects across platforms, all in one place. One profile, one application process, more opportunities.