Skip to content
← Back to explorer

HFEPX Hub

CS.SE + Long Horizon Papers

Updated from current HFEPX corpus (Apr 9, 2026). 12 papers are grouped in this hub page.

Read Full Context

Updated from current HFEPX corpus (Apr 9, 2026). 12 papers are grouped in this hub page. Common evaluation modes: Automatic Metrics. Most common rater population: Domain Experts. Common annotation unit: Trajectory. Frequently cited benchmark: SWE-bench. Common metric signal: cost. 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: 12 Last published: Apr 8, 2026 Global RSS Tag RSS
Cs.SELong Horizon

Researcher Quick Triage

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

All Sampled Papers (12) Replication-Ready Only (2)

High-Signal Coverage

100.0%

12 / 12 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.
  • 0 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

  • 16.7% of papers report explicit human-feedback signals, led by red-team protocols.
  • automatic metrics appears in 50% of papers in this hub.
  • SWE-bench is a recurring benchmark anchor for cross-paper comparisons in this page.

Protocol Takeaways

  • Quality-control reporting is sparse in this slice; prioritize papers with explicit calibration or adjudication steps.
  • Rater context is mostly domain experts, and annotation is commonly trajectory-level annotation; use this to scope replication staffing.
  • Stratify by benchmark (SWE-bench vs SWE-bench Verified) before comparing methods.

Benchmark Interpretation

  • SWE-bench appears in 33.3% of hub papers (2/12); use this cohort for benchmark-matched comparisons.
  • SWE-bench Verified appears in 33.3% of hub papers (2/12); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • cost is reported in 33.3% of hub papers (2/12); compare with a secondary metric before ranking methods.
  • pass@1 is reported in 33.3% of hub papers (2/12); compare with a secondary metric before ranking methods.
Researcher Checklist (Expanded)

Researcher Checklist

  • Gap: Papers with explicit human feedback

    Coverage is a replication risk (16.7% vs 45% target).

  • Gap: Papers reporting quality controls

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

  • Strong: Papers naming benchmarks/datasets

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

  • Strong: Papers naming evaluation metrics

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

  • Gap: Papers with known rater population

    Coverage is a replication risk (16.7% vs 35% target).

  • Strong: Papers with known annotation unit

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

Strengths

  • Most papers provide measurable evaluation context (50% benchmarks, 100% metrics).
  • Agentic evaluation appears in 100% of papers.

Known Gaps

  • Only 0% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (16.7% coverage).

Suggested Next Analyses

  • Stratify by benchmark (SWE-bench vs SWE-bench Verified) before comparing methods.
  • Track metric sensitivity by reporting both cost and pass@1.
Recommended Queries (Expanded)

Recommended Queries

Benchmark Slice: SWE-bench Metric Slice: cost 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
TraceSafe: A Systematic Assessment of LLM Guardrails on Multi-Step Tool-Calling Trajectories

Apr 8, 2026

 Yes
Red Team
Automatic Metrics Tracesafe Bench Accuracy Not Reported
SWE-Protégé: Learning to Selectively Collaborate With an Expert Unlocks Small Language Models as Software Engineering Agents

Feb 25, 2026

 No
Not Reported
Automatic Metrics SWE Bench , SWE Bench Verified Pass@1 , Latency Not Reported
Structurally Aligned Subtask-Level Memory for Software Engineering Agents

Feb 25, 2026

 No
Not Reported
Automatic Metrics SWE Bench Not Reported Not Reported
The Evolution of Tool Use in LLM Agents: From Single-Tool Call to Multi-Tool Orchestration

Mar 24, 2026

 No
Not Reported
Automatic Metrics Not Reported Cost Not Reported
SkillCraft: Can LLM Agents Learn to Use Tools Skillfully?

Feb 28, 2026

 No
Not Reported
Automatic Metrics Not Reported Success rate Not Reported
ToolMATH: A Math Tool Benchmark for Realistic Long-Horizon Multi-Tool Reasoning

Feb 24, 2026

 No
Not Reported
Automatic Metrics Not Reported Coherence Not Reported
WebTestBench: Evaluating Computer-Use Agents towards End-to-End Automated Web Testing

Mar 26, 2026

 No
Not Reported
Not Reported Webtestbench Not Reported Not Reported
SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

Mar 25, 2026

 No
Not Reported
Not Reported Slopcodebench Not Reported Not Reported
Beyond Local Code Optimization: Multi-Agent Reasoning for Software System Optimization

Mar 16, 2026

 No
Not Reported
Not Reported Not Reported Throughput Not Reported
ReflexiCoder: Teaching Large Language Models to Self-Reflect on Generated Code and Self-Correct It via Reinforcement Learning

Mar 6, 2026

 No
Not Reported
Not Reported LiveCodeBench , HumanEval+ Not Reported Not Reported
SWE-Master: Unleashing the Potential of Software Engineering Agents via Post-Training

Feb 3, 2026

 No
Not Reported
Not Reported SWE Bench , SWE Bench Verified Not Reported Not Reported
Process-Centric Analysis of Agentic Software Systems

Dec 2, 2025

 No
Not Reported
Not Reported SWE Bench Not Reported Not Reported

Protocol Diff (Top Papers)

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

Signal TraceSafe: A Systematic Assessment of LLM Guardrail… SWE-Protégé: Learning to Selectively Collaborate Wi… Structurally Aligned Subtask-Level Memory for Softw…
Human Feedback Red TeamNot reportedNot reported
Evaluation Modes Automatic MetricsAutomatic MetricsAutomatic Metrics
Benchmarks Tracesafe BenchSWE Bench, SWE Bench VerifiedSWE Bench
Metrics AccuracyPass@1, LatencyNot reported
Quality Controls Not reportedNot reportedNot reported
Rater Population UnknownDomain ExpertsUnknown
Annotation Unit TrajectoryUnknownTrajectory
Suggested Reading Order (Extended)

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

Suggested Reading Order

  1. TraceSafe: A Systematic Assessment of LLM Guardrails on Multi-Step Tool-Calling Trajectories

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics + red-team protocols. Focus: Tracesafe-Bench / accuracy. Abstract: As large language models (LLMs) evolve from static.

  2. The Evolution of Tool Use in LLM Agents: From Single-Tool Call to Multi-Tool Orchestration

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics. Focus: cost. Abstract: Tool use enables large language models (LLMs) to access external information, invoke software.

  3. SkillCraft: Can LLM Agents Learn to Use Tools Skillfully?

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics. Focus: success rate. Abstract: We further propose a lightweight evaluation protocol that enables agents to auto-compose.

  4. SWE-Protégé: Learning to Selectively Collaborate With an Expert Unlocks Small Language Models as Software Engineering Agents

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: automatic metrics. Focus: SWE-bench / pass@1. Abstract: Small language models (SLMs) offer compelling advantages in cost, latency,.

  5. Structurally Aligned Subtask-Level Memory for Software Engineering Agents

    Adds automatic metrics for broader protocol coverage within this hub. Signals: automatic metrics. Focus: SWE-bench. Abstract: Large Language Models (LLMs) have demonstrated significant potential as autonomous software engineering.

  6. ToolMATH: A Math Tool Benchmark for Realistic Long-Horizon Multi-Tool Reasoning

    Adds automatic metrics for broader protocol coverage within this hub. Signals: automatic metrics. Focus: coherence. Abstract: We introduce \ToolMATH, a math-grounded benchmark that evaluates tool-augmented language models in.

  7. WebTestBench: Evaluating Computer-Use Agents towards End-to-End Automated Web Testing

    Adds evaluation protocol evidence for broader protocol coverage within this hub. Focus: Webtestbench. Abstract: The emergence of Large Language Models (LLMs) has catalyzed a paradigm shift in programming,.

  8. SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

    Adds evaluation protocol evidence for broader protocol coverage within this hub. Focus: Slopcodebench. Abstract: Software development is iterative, yet agentic coding benchmarks overwhelmingly evaluate single-shot solutions against complete.

Known Limitations

Known Limitations

  • Only 0% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (16.7% 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

  • Red Team (1)

Evaluation Modes

  • Automatic Metrics (6)

Top Benchmarks

  • SWE Bench (2)
  • SWE Bench Verified (2)
  • Tracesafe Bench (1)

Top Metrics

  • Cost (2)
  • Pass@1 (2)
  • Accuracy (1)
  • Coherence (1)

Rater Population Mix

  • Domain Experts (1)

Quality Controls

Coverage diagnostics (sample-based): human-feedback 8.3% · benchmarks 66.7% · metrics 50.0% · quality controls 0.0%.

Top Papers

Related Hubs

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

Post a Job Get a Quote