Skip to content
← Back to explorer

HFEPX Hub

Tool Use + Coding (Last 120 Days)

Updated from current HFEPX corpus (Apr 12, 2026). 13 papers are grouped in this hub page.

Read Full Context

Updated from current HFEPX corpus (Apr 12, 2026). 13 papers are grouped in this hub page. Common evaluation modes: Automatic Metrics, Simulation Env. Most common rater population: Domain Experts. Common annotation unit: Trajectory. Frequently cited benchmark: agent-diff-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 Mar 25, 2026.

Papers: 13 Last published: Mar 25, 2026 Global RSS Tag RSS
Tool UseCodingLast 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%

13 / 13 sampled papers are not low-signal flagged.

Replication-Ready Set

3

Benchmark + metric + eval mode explicitly present.

Judge/Human Comparability

0

Papers containing both `human_eval` and `llm_as_judge`.

  • 3 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: 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

  • 38.5% of papers report explicit human-feedback signals, led by pairwise preferences.
  • automatic metrics appears in 69.2% of papers in this hub.
  • agent-diff-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 (agent-diff-bench vs BrowseComp) before comparing methods.

Benchmark Interpretation

  • agent-diff-bench appears in 7.7% of hub papers (1/13); use this cohort for benchmark-matched comparisons.
  • BrowseComp appears in 7.7% of hub papers (1/13); use this cohort for benchmark-matched comparisons.

Metric Interpretation

  • cost is reported in 38.5% of hub papers (5/13); compare with a secondary metric before ranking methods.
  • accuracy is reported in 23.1% of hub papers (3/13); compare with a secondary metric before ranking methods.
Researcher Checklist (Expanded)

Researcher Checklist

  • Moderate: Papers with explicit human feedback

    Coverage is usable but incomplete (38.5% 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 (38.5% vs 35% target).

  • Strong: Papers naming evaluation metrics

    Coverage is strong (76.9% vs 35% target).

  • Gap: Papers with known rater population

    Coverage is a replication risk (15.4% vs 35% target).

  • Gap: Papers with known annotation unit

    Coverage is a replication risk (15.4% vs 35% target).

Strengths

  • Most papers provide measurable evaluation context (38.5% benchmarks, 76.9% 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 (15.4% coverage).
  • Annotation unit is under-specified (15.4% coverage).

Suggested Next Analyses

  • Stratify by benchmark (agent-diff-bench vs BrowseComp) before comparing methods.
  • Track metric sensitivity by reporting both cost and accuracy.
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
Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters

Feb 11, 2026

Yes Not Reported LiveCodeBench , BrowseComp Latency , Cost 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
ToolFlood: Beyond Selection -- Hiding Valid Tools from LLM Agents via Semantic Covering

Mar 14, 2026

No
Not Reported
Automatic Metrics ToolBench Success rate , Jailbreak success rate Not Reported
Sabiá-4 Technical Report

Mar 10, 2026

Yes Automatic Metrics Not Reported Accuracy , Cost Not Reported
Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception

Feb 12, 2026

No
Not Reported
Automatic Metrics Zoombench Latency Not Reported
Agent-Diff: Benchmarking LLM Agents on Enterprise API Tasks via Code Execution with State-Diff-Based Evaluation

Feb 11, 2026

No
Not Reported
Automatic Metrics Agent Diff Bench Task success Not Reported
The Ultimate Tutorial for AI-driven Scale Development in Generative Psychometrics: Releasing AIGENIE from its Bottle

Mar 30, 2026

Yes Not Reported Not Reported Not Reported Not Reported
Tucano 2 Cool: Better Open Source LLMs for Portuguese

Mar 3, 2026

Yes Not Reported Not Reported 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
The Detection-Extraction Gap: Models Know the Answer Before They Can Say It

Apr 8, 2026

No
Not Reported
Automatic Metrics Not Reported Accuracy , Cost Not Reported
AgentGL: Towards Agentic Graph Learning with LLMs via Reinforcement Learning

Apr 7, 2026

No
Not Reported
Automatic Metrics Not Reported Accuracy Not Reported
REDSearcher: A Scalable and Cost-Efficient Framework for Long-Horizon Search Agents

Feb 15, 2026

No
Not Reported
Automatic Metrics Not Reported Recall , Cost Not Reported

Protocol Diff (Top Papers)

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

Signal Step 3.5 Flash: Open Frontier-Level Intelligence wi… VehicleMemBench: An Executable Benchmark for Multi-… ToolFlood: Beyond Selection -- Hiding Valid Tools f…
Human Feedback Pairwise PreferencePairwise PreferenceNot reported
Evaluation Modes Not reportedSimulation EnvAutomatic Metrics
Benchmarks LiveCodeBench, BrowseCompVehiclemembenchToolBench
Metrics Latency, CostNot reportedSuccess rate, Jailbreak success rate
Quality Controls Not reportedNot reportedNot reported
Rater Population Domain ExpertsUnknownUnknown
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. The Detection-Extraction Gap: Models Know the Answer Before They Can Say It

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics. Focus: accuracy. Abstract: Modern reasoning models continue generating long after the answer is already determined.

  2. AgentGL: Towards Agentic Graph Learning with LLMs via Reinforcement Learning

    Start here for detailed protocol reporting and quality-control evidence. Signals: automatic metrics. Focus: accuracy. Abstract: Large Language Models (LLMs) increasingly rely on agentic capabilities-iterative retrieval, tool use, and.

  3. The Ultimate Tutorial for AI-driven Scale Development in Generative Psychometrics: Releasing AIGENIE from its Bottle

    Start here for detailed protocol reporting and quality-control evidence. Signals: critique/edit feedback. Abstract: Psychological scale development has traditionally required extensive expert involvement, iterative revision, and large-scale pilot testing.

  4. 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.

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

    Include a human-eval paper to calibrate against judge-based evaluation settings. Signals: pairwise preferences. Focus: LiveCodeBench / latency. Abstract: To reach frontier-level intelligence, we design a scalable reinforcement learning.

  6. Sabiá-4 Technical Report

    Adds automatic metrics with pairwise preferences for broader protocol coverage within this hub. Signals: automatic metrics + pairwise preferences. Focus: accuracy. Abstract: The models were developed through a.

  7. Tucano 2 Cool: Better Open Source LLMs for Portuguese

    Adds evaluation protocol evidence with pairwise preferences for broader protocol coverage within this hub. Signals: pairwise preferences. Abstract: Following our previous works, we now extend our dataset, GigaVerbo-v2,.

  8. ToolFlood: Beyond Selection -- Hiding Valid Tools from LLM Agents via Semantic Covering

    Adds automatic metrics for broader protocol coverage within this hub. Signals: automatic metrics. Focus: ToolBench / success rate. Abstract: Large Language Model (LLM) agents increasingly use external tools.

Known Limitations

Known Limitations

  • Only 0% of papers report quality controls; prioritize calibration/adjudication evidence.
  • Rater population is under-specified (15.4% 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 (4)
  • Critique Edit (1)

Evaluation Modes

  • Automatic Metrics (9)
  • Simulation Env (1)

Top Benchmarks

  • Agent Diff Bench (1)
  • BrowseComp (1)
  • Imo Answerbench (1)
  • LiveCodeBench (1)

Top Metrics

  • Cost (5)
  • Accuracy (3)
  • Latency (2)
  • F1 (1)

Rater Population Mix

  • Domain Experts (2)

Quality Controls

Coverage diagnostics (sample-based): human-feedback 38.5% · benchmarks 38.5% · metrics 76.9% · quality controls 0.0%.

Top Papers

Related Hubs

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