Event

NeurIPS 2026

NeurIPS 2026

NeurIPS 2026

Abstract image

The hidden infrastructure behind world-class AI models

The hidden infrastructure behind world-class AI models

The hidden infrastructure behind world-class AI models

Overview

Centific at NeurIPS 2026

Bring your toughest AI testing question

What would an AI agent have to prove before you’d trust it to act on a customer’s behalf? Centific’s Applied AI Research team is bringing eight papers to the NeurIPS 2026 workshops. Come meet the researchers, compare notes on how you test your own systems, and ask what they’d put in a benchmark for the work you need AI to do.

Sidney, Australia

Location (Main)

Location

Atlanta, US

Location (Satellite)

Date

Paris, France

Location (Satellite)

Date

Find us

Where AI fails and how to test for it

Where AI fails and how to test for it

At NeurIPS, our researchers will discuss questions like, when should a phone assistant hold back? Who benefits when AI agents close a deal? Our studies also test voice agents on calls, probe gender bias in speech models, and analyze hours-long recordings. We built benchmark tasks for finance, tested how multilingual answer keys change scores, and graded healthcare AI on complete deliverables.

At NeurIPS, our researchers will discuss questions like, when should a phone assistant hold back? Who benefits when AI agents close a deal? Our studies also test voice agents on calls, probe gender bias in speech models, and analyze hours-long recordings. We built benchmark tasks for finance, tested how multilingual answer keys change scores, and graded healthcare AI on complete deliverables.

Research papers and publications

Centific papers presentation sessions

Centific papers presentation sessions

STRATA: Automated Construction of Professional Tasks and Rubrics for Agent Evaluation

Location

Sydney

Time

Fri, Dec 11

TBA

Can voice agents help you get through the day?

Location

Sydney - room MR C4.11

Time

Sat, Dec 12

Poster Session 1, 12:15 PM

Same Output, Different Gold: Measuring How Reference Choice Moves a Multilingual Benchmark Score

Location

Sydney - room MR C2.2 & C2.3

Time

Sat, Dec 12

TBA

Voice or Stereotype? Disentangling Acoustic and Content-Based Gender in Speech-to-Speech Models

Location

Sydney - room MR C2.2 & C2.3

Time

Sat, Dec 12

TBA

Evaluating deliverables, not answers, in health occupations

Location

Sydney

Time

Sat, Dec 12

Poster session, 3:50–4:40 PM

AudioLoop: Program-Guided Iterative Reasoning Over Long-Form Speech

Location

Atlanta - room C105–C106

Time

Sat, Dec 12

TBA

Closing the Deal Is Not Enough: A Trajectory-Level Behavioral Rubric for Interactive Agents

Location

Atlanta

Time

Sun, Dec 13

TBA

PAVE: Device-Verified Evaluation of Action and Restraint in Personal Assistants

Location

Paris

Time

Sun, Dec 13

TBA

Open Positions

We're hiring

We're hiring

come and join our research team

come and join our research team

Discover

More ways to connect

Connect with Centific

Stay ahead of what’s next

Stay ahead

Updates from the frontier of AI data.

Receive updates on platform improvements, new workflows, evaluation capabilities, data quality enhancements, and best practices for enterprise AI teams.

By proceeding, you agree to our Terms of Use and Privacy Policy