Skip to content
← Back to explorer

HFEPX Hub

Coding + Pairwise Preference (Last 120 Days)

Updated from current HFEPX corpus (Apr 17, 2026). 34 papers are grouped in this hub page.

Read Full Context

Updated from current HFEPX corpus (Apr 17, 2026). 34 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: APPS. 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 Mar 25, 2026.

Papers: 34 Last published: Mar 25, 2026 Global RSS Tag RSS
CodingPairwise PreferenceLast 120d

Researcher Quick Triage

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

High-Signal Coverage

100.0%

34 / 34 sampled papers are not low-signal flagged.

Replication-Ready Set

4

Benchmark + metric + eval mode explicitly present.

Judge/Human Comparability

0

Papers containing both `human_eval` and `llm_as_judge`.

  • 4 papers are replication-ready (benchmark + metric + explicit evaluation mode).
  • 0 papers support judge-vs-human agreement analysis.
  • 1 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 47.1% of papers in this hub.
  • APPS is a recurring benchmark anchor for cross-paper comparisons in this page.

Protocol Takeaways

  • Most common quality-control signal is rater calibration (2.9% 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

  • APPS appears in 5.9% of hub papers (2/34); use this cohort for benchmark-matched comparisons.
  • LiveCodeBench appears in 5.9% of hub papers (2/34); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • accuracy is reported in 23.5% of hub papers (8/34); compare with a secondary metric before ranking methods.
  • cost is reported in 17.6% of hub papers (6/34); 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 (2.9% vs 30% target).

  • Moderate: Papers naming benchmarks/datasets

    Coverage is usable but incomplete (26.5% vs 35% target).

  • Strong: Papers naming evaluation metrics

    Coverage is strong (50% vs 35% target).

  • Gap: Papers with known rater population

    Coverage is a replication risk (11.8% vs 35% target).

  • Strong: Papers with known annotation unit

    Coverage is strong (35.3% 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 2.9% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (11.8% coverage).
  • LLM-as-judge appears without enough inter-annotator agreement reporting.

Suggested Next Analyses

  • Compare papers that report both human_eval and llm_as_judge to quantify judge-human agreement drift.
  • Stratify by benchmark (APPS vs LiveCodeBench) before comparing methods.
  • Track metric sensitivity by reporting both accuracy and cost.
  • Add inter-annotator agreement checks when reproducing these protocols.
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
Do Phone-Use Agents Respect Your Privacy?

Apr 1, 2026

Yes Automatic Metrics APPS , Myphonebench Task success Not Reported
CausalRM: Causal-Theoretic Reward Modeling for RLHF from Observational User Feedbacks

Mar 19, 2026

Yes Automatic Metrics Harmbench Not Reported Not Reported
Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters

Feb 11, 2026

Yes Not Reported LiveCodeBench , BrowseComp Not Reported Not 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
RewardUQ: A Unified Framework for Uncertainty-Aware Reward Models

Feb 27, 2026

Yes Automatic Metrics Not Reported Accuracy Calibration
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
IntelliAsk: Learning to Ask High-Quality Research Questions via RLVR

Jan 23, 2026

Yes Human Eval Writingbench Not Reported Not Reported
ChartEditBench: Evaluating Grounded Multi-Turn Chart Editing in Multimodal Language Models

Feb 17, 2026

Yes Automatic Metrics Charteditbench Not Reported Not Reported
FEAST: Fully Connected Expressive Attention for Spatial Transcriptomics

Mar 26, 2026

Yes Not Reported Not Reported Not Reported Not Reported
IslamicMMLU: A Benchmark for Evaluating LLMs on Islamic Knowledge

Mar 24, 2026

Yes Automatic Metrics Not Reported Accuracy Not Reported
From Oracle to Noisy Context: Mitigating Contextual Exposure Bias in Speech-LLMs

Mar 25, 2026

Yes Not Reported Not Reported Wer , Jailbreak success rate Not Reported

Protocol Diff (Top Papers)

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

Signal Do Phone-Use Agents Respect Your Privacy? CausalRM: Causal-Theoretic Reward Modeling for RLHF… Step 3.5 Flash: Open Frontier-Level Intelligence wi…
Human Feedback Pairwise PreferencePairwise PreferencePairwise Preference
Evaluation Modes Automatic MetricsAutomatic MetricsNot reported
Benchmarks APPS, MyphonebenchHarmbenchLiveCodeBench, BrowseComp
Metrics Task successNot reportedNot reported
Quality Controls Not reportedNot reportedNot reported
Rater Population UnknownUnknownDomain Experts
Annotation Unit UnknownUnknownUnknown
Suggested Reading Order (Extended)

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

Suggested Reading Order

  1. Do Phone-Use Agents Respect Your Privacy?

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics + pairwise preferences. Focus: APPS / task success. Abstract: Across five frontier models on 10 mobile.

  2. FEAST: Fully Connected Expressive Attention for Spatial Transcriptomics

    Start here for detailed protocol reporting and quality-control evidence. Signals: pairwise preferences. Focus: cost. Abstract: To address this, we propose FEAST (Fully connected Expressive Attention for Spatial Transcriptomics),.

  3. Comparing Developer and LLM Biases in Code Evaluation

    Start here for detailed protocol reporting and quality-control evidence. Signals: pairwise preferences. Abstract: As LLMs are increasingly used as judges in code applications, they should be evaluated in.

  4. IntelliAsk: Learning to Ask High-Quality Research Questions via RLVR

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: human evaluation + pairwise preferences. Focus: Writingbench. Abstract: To address this gap, we curate a high-quality dataset.

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

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: simulation environments + pairwise preferences. Focus: Vehiclemembench. Abstract: This evolution requires agents to continuously model multi-user preferences.

  6. WebCoderBench: Benchmarking Web Application Generation with Comprehensive and Interpretable Evaluation Metrics

    Include an LLM-as-judge paper to test judge design and agreement assumptions. Signals: LLM-as-judge + pairwise preferences. Focus: LMSYS Chatbot Arena. Abstract: WebCoderBench provides 24 fine-grained evaluation metrics across.

  7. Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters

    Adds evaluation protocol evidence with pairwise preferences for broader protocol coverage within this hub. Signals: pairwise preferences. Focus: LiveCodeBench / latency. Abstract: To reach frontier-level intelligence, we design.

  8. Decoupling Strategy and Execution in Task-Focused Dialogue via Goal-Oriented Preference Optimization

    Adds automatic metrics with pairwise preferences for broader protocol coverage within this hub. Signals: automatic metrics + pairwise preferences. Focus: task success. Abstract: Large language models show potential.

Known Limitations

Known Limitations

  • Only 2.9% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (11.8% 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 (34)
  • Expert Verification (2)
  • Critique Edit (1)
  • Rubric Rating (1)

Evaluation Modes

  • Automatic Metrics (16)
  • Human Eval (1)
  • Llm As Judge (1)
  • Simulation Env (1)

Top Benchmarks

  • APPS (2)
  • LiveCodeBench (2)
  • AIME (1)
  • BrowseComp (1)

Top Metrics

  • Accuracy (8)
  • Cost (6)
  • Latency (2)
  • Task success (2)

Rater Population Mix

  • Domain Experts (3)
  • Mixed (1)

Quality Controls

  • Calibration (1)
Coverage diagnostics (sample-based): human-feedback 100.0% · benchmarks 26.5% · metrics 50.0% · quality controls 2.9%.

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.