DH{}DataHunter

Dataset de Hugging Face · Salesforce

LALM Judge Validation on Full-Duplex Voice Agents

Abrir en Hugging Face
Licencia
CC BY 4.0uso comercial permitidoTexto de la licencia
Esta licencia permite el uso comercial. Normalmente hay que citar al autor; consulta la licencia para ver las condiciones exactas.
Formatos
CSV, CFF, MD, PNG, PY
Tamaño
981 kB
Actualizado
23 jul 2026

Descripción

LALM Judge Validation on Full-Duplex Voice Agents Companion dataset for the paper A Reliability Assessment of LALM Audio Judges for Full-Duplex Voice Agents. This repository contains the anonymised ratings, adversarial-defect recall tables, JSON schemas, and analysis scripts used to produce every headline number, table, and figure in that paper. Summary 209 rated stereo sessions: 152 full-duplex agent-client conversations across 13 accent-and-condition strata…

Descripción de Hugging Face, abreviada.

Archivos (27)

  • CFFCITATION.cff1004 B
  • MDCODE_OF_CONDUCT.md5 kB
  • MDCONTRIBUTING.md2,3 kB
  • —LICENSE12 kB
  • MDSECURITY.md452 B
  • CSVdata/cross_model_comparison.csv1,2 kB
  • CSVdata/d_personality_vs_codec.csv625 B
  • CSVdata/e_ceiling_collapse.csv1,6 kB
  • CSVdata/full_per_session.csv133 kB
  • CSVdata/h1_per_dim.csv1,2 kB
  • CSVdata/h1_per_dim_gemini31propreview.csv640 B
  • CSVdata/h1_per_dim_gemini35flash.csv639 B
  • CSVdata/h1_per_dim_natural.csv516 B
  • CSVdata/h2_per_config_per_dim.csv6 kB
  • CSVdata/h3_defect_recall_3rater.csv2,7 kB
  • CSVdata/h3_defect_recall_with_ci.csv4,3 kB
  • CSVdata/lalm_as_4th_rater.csv876 B
  • CSVdata/rho_hh_ci.csv528 B
  • PNGfigures/fig1_three_regime_taxonomy.png150 kB
  • PNGfigures/fig2_lalm_as_4th_rater.png75 kB
  • PNGfigures/fig3_defect_detection_forest.png340 kB
  • PNGfigures/fig4_personality_vs_codec.png112 kB
  • PNGfigures/fig5_ceiling_partition.png119 kB
  • MDscripts/README.md1,4 kB
  • PYscripts/compute_agreement.py4 kB
  • PYscripts/compute_krippendorff.py2,7 kB
  • PYscripts/compute_newcombe_wilson.py3,9 kB

DataHunter no guarda ni aloja estos archivos: cada enlace lleva a la fuente original. Revisa allí la licencia antes de usar los datos.

Buscar otros datasets