Slides: The Art & Science of Benchmarking Agents — Vincent Chen, Snorkel AI
Source Video
The Art & Science of Benchmarking Agents — Vincent Chen, Snorkel AI
Relationship To World's Fair 2026
These slides are extracted from a public AI Engineer YouTube video connected to World's Fair 2026. Speaker-matched clips are supporting context unless later confirmed as exact session recordings; official livestream recordings are day-level/event-level source material.
Related Scheduled Sessions
- No individual scheduled session mapping has been assigned yet; treat this as an event livestream deck.
Extracted Slides
OCR text:
Snorkel
The Art & Science
ofBenchmarking
Agents
OCR text:
@ Snorkel te Frontier al bata tab
Alex Ratner Sermareners FOES add Aapiied Al Engrimers truce QOD + 2 reer acer papers pottanis
ve ae PAyetty and Traits eat tutions woraing on dale on dats Cerne Al gent feaipasnion aeahey
a*% Meantane bs Contre AL st ewiling, pital tabs — ay
ab tf AM pte tetas. at 5 é; o
a, jee 8 C60 4 O8in. Be © A eneira demtinaes Oace
© item sucetae
* bd ooo
Chris Ri Star Se CUTA = 2 8 SAM Lge stan ot Lik i IT SeD
bf bd pane ; © SOLA pau eet eS Jaren this andetitandeg and
ra * ; crm Bune gun Yale Eating Spouse
@e ene ee Aamards at Newt, KL CLR UAL VLO8, and more
Icepehitee Weta, A
a; TEER ES! erpeigcarnemenpennneeeTE ee Comuthored / contributed te 20+ benchenarhs
Frod Sala “este, tabed we bencbnan | eae ae © tyenicany Bhar aeemgnapeaa
Oo Marne ace y aie beer arent ata rege peteret Sor eo teem cect atm eters
5 SALERUIRRES EE agesbacanlh Summsens es ie vaydticben peen
lerigneney Hoes ERAS SOiERNE ra sIRKEMEN, sas
ee 5 ‘ 2 AUER Fen eeeerrre
= NH Ae a ations ated er mee ne gen pawn
Ie eSe asco atalys aed tere eg 2 thy ware,
8 Mateo Can rarer arccon
Bret a tea teats
ee
Fd
OCR text:
an Benchmarks
x .
5g bd
Sd ra bd
A S3M+ commitment to
open benchmarks
PRCmSTaL dC) oes al a Fa a ans togetherar FACTORY harbor (on ea o
| Al Engineer |
EUROPE
i Pe a
Asad
OCR text:
Difficulty / Model Headroom -
Benchmark is unsaturated. 7
a * -
* ad ik ca
rs * ARC-AGI-3 er ae
5g Sd
an el Shipped with <1% model score (100% of
raster geretsolvableby homans!
Engineering the future of Al
;
aay
OCR text:
Lasting benchmarks push
od the frontier
| Al Engineer |
EUROPE
; FT
Sa
OCR text:
Snorkelm Fore a
Propose a benchmark
benchmarks.snorkelal
：@vincentsunnchen
gineer
OCR text:
| Al Engineer |
aU elas
AI.ENGINEER
Slide-Derived Subjects To Review
Subject extraction uses video title, related session titles/descriptions, transcript context, and OCR text when available. OCR is best-effort and should be reviewed against the embedded slide images.