Skip to content
← Back to explorer

HFEPX Hub

CS.HC Human Feedback And Eval Papers

Updated from current HFEPX corpus (Feb 27, 2026). 35 papers are grouped in this hub page. Common evaluation modes: Automatic Metrics, Simulation Env. Most common rater population: Domain Experts. Common annotation unit: Multi Dim Rubric. Frequently cited benchmark: Retrieval. 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 26, 2026.

Papers: 35 Last published: Feb 26, 2026 Global RSS
Cs.HC

Research Narrative

Grounded narrative Model: deterministic-grounded Source: persisted

Updated from current HFEPX corpus (Feb 27, 2026). This page tracks 35 papers for CS.HC Human Feedback And Eval Papers. Dominant protocol signals include automatic metrics, simulation environments, human evaluation, with frequent benchmark focus on Retrieval and metric focus on accuracy, cost. Use the grounded sections below to prioritize reproducible protocol choices, benchmark-matched comparisons, and judge-vs-human evaluation checks.

Why This Matters For Eval Research

Protocol Takeaways

Benchmark Interpretation

  • Retrieval appears in 2.9% of hub papers (1/35); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • accuracy is reported in 14.3% of hub papers (5/35); compare with a secondary metric before ranking methods.
  • cost is reported in 8.6% of hub papers (3/35); compare with a secondary metric before ranking methods.

Researcher Checklist

  • Tighten coverage on Papers with explicit human feedback. Coverage is usable but incomplete (34.3% vs 45% target).
  • Close gap on Papers reporting quality controls. Coverage is a replication risk (0% vs 30% target).
  • Close gap on Papers naming benchmarks/datasets. Coverage is a replication risk (2.9% vs 35% target).
  • Tighten coverage on Papers naming evaluation metrics. Coverage is usable but incomplete (31.4% vs 35% target).
  • Tighten coverage on Papers with known rater population. Coverage is usable but incomplete (25.7% vs 35% target).
  • Close gap on Papers with known annotation unit. Coverage is a replication risk (11.4% vs 35% target).

Papers with explicit human feedback

Coverage is usable but incomplete (34.3% vs 45% target).

Papers reporting quality controls

Coverage is a replication risk (0% vs 30% target).

Papers naming benchmarks/datasets

Coverage is a replication risk (2.9% vs 35% target).

Papers naming evaluation metrics

Coverage is usable but incomplete (31.4% vs 35% target).

Papers with known rater population

Coverage is usable but incomplete (25.7% vs 35% target).

Papers with known annotation unit

Coverage is a replication risk (11.4% vs 35% target).

Suggested Reading Order

  1. 1. LLM Novice Uplift on Dual-Use, In Silico Biology Tasks

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

  2. 2. TherapyProbe: Generating Design Knowledge for Relational Safety in Mental Health Chatbots Through Adversarial Simulation

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

  3. 3. Dynamic Personality Adaptation in Large Language Models via State Machines

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

  4. 4. When AI Writes, Whose Voice Remains? Quantifying Cultural Marker Erasure Across World English Varieties in Large Language Models

    Adds automatic metrics for broader coverage within this hub.

  5. 5. Exploring Human-Machine Coexistence in Symmetrical Reality

    Adds automatic metrics for broader coverage within this hub.

  6. 6. Evaluating the Usage of African-American Vernacular English in Large Language Models

    Adds automatic metrics for broader coverage within this hub.

  7. 7. SparkMe: Adaptive Semi-Structured Interviewing for Qualitative Insight Discovery

    Adds automatic metrics with expert verification for broader coverage within this hub.

  8. 8. "Are You Sure?": An Empirical Study of Human Perception Vulnerability in LLM-Driven Agentic Systems

    Adds automatic metrics with expert verification for broader coverage within this hub.

Known Limitations

  • Only 0% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Annotation unit is under-specified (11.4% coverage).
  • Narrative synthesis is grounded in metadata and abstracts only; full-paper implementation details are not parsed.

Research Utility Links

human_eval vs automatic_metrics

both=0, left_only=1, right_only=28

0 papers use both Human Eval and Automatic Metrics.

automatic_metrics vs simulation_env

both=0, left_only=28, right_only=6

0 papers use both Automatic Metrics and Simulation Env.

simulation_env vs human_eval

both=0, left_only=6, right_only=1

0 papers use both Simulation Env and Human Eval.

Benchmark Brief

Retrieval

Coverage: 1 papers (2.9%)

1 papers (2.9%) mention Retrieval.

Examples: Toward Safe and Human-Aligned Game Conversational Recommendation via Multi-Agent Decomposition

Metric Brief

accuracy

Coverage: 5 papers (14.3%)

5 papers (14.3%) mention accuracy.

Examples: Modeling Distinct Human Interaction in Web Agents , What Do LLMs Associate with Your Name? A Human-Centered Black-Box Audit of Personal Data , How much does context affect the accuracy of AI health advice?

Metric Brief

cost

Coverage: 3 papers (8.6%)

3 papers (8.6%) mention cost.

Examples: When AI Writes, Whose Voice Remains? Quantifying Cultural Marker Erasure Across World English Varieties in Large Language Models , SparkMe: Adaptive Semi-Structured Interviewing for Qualitative Insight Discovery , A Scalable Framework for Evaluating Health Language Models

Metric Brief

agreement

Coverage: 2 papers (5.7%)

2 papers (5.7%) mention agreement.

Examples: Through the Judge's Eyes: Inferred Thinking Traces Improve Reliability of LLM Raters , A Scalable Framework for Evaluating Health Language Models

Most Cited In This Hub

Fast path to methods with the strongest citation traction in this scope.

Papers: LLM Novice Uplift on Dual-Use, In Silico Biology Tasks , TherapyProbe: Generating Design Knowledge for Relational Safety in Mental Health Chatbots Through Adversarial Simulation , Dynamic Personality Adaptation in Large Language Models via State Machines

Most Recent

Fast path to latest protocol changes and newly published evaluation setups.

Papers: LLM Novice Uplift on Dual-Use, In Silico Biology Tasks , TherapyProbe: Generating Design Knowledge for Relational Safety in Mental Health Chatbots Through Adversarial Simulation , Dynamic Personality Adaptation in Large Language Models via State Machines

Best Protocol Detail

Papers with explicit rater/unit metadata and quality-control signals for reproducibility.

Papers: LLM Novice Uplift on Dual-Use, In Silico Biology Tasks , TherapyProbe: Generating Design Knowledge for Relational Safety in Mental Health Chatbots Through Adversarial Simulation , Dynamic Personality Adaptation in Large Language Models via State Machines

Top Papers

Related Hubs