@jn
Research Engineer @ Apart Research
https://www.linkedin.com/in/jonathan-ng-7061a3162/$0 in pending offers
Jonathan has extensive experience in organizing and leading AI safety initiatives. He:
Directed AI Safety Fellowships at both NUS and NTU, leading two independent teams of three student organizers each.
Graduated from the pilot cohort of ARENA (a leading AI safety training program).
Completed the MATS (ML Alignment & Theory Scholars) Spring 2023 program.
Authored notable projects including MACHIAVELLI (ICML 2023) (a benchmark for studying deceptive AI behavior) and cybercapabilities.org (AAAI ‘25 Workshop) (an agentic cyber capabilities benchmark)
Jonathan Ng
19 days ago
@RyanKidd Just following up on this. Yanni responded to my email, suggesting to:
1. Acknowledge where we have drawn from the TARA grant application
2. Change the name, purely to keep differentiation between programs
I’ve updated the proposal to explicitly acknowledge the TARA team's contribution and changed the program name to RECAP to maintain clear differentiation. Going forward, I'll be more mindful of attribution and derivation in my work.
Thanks again for calling this out.
Jonathan Ng
24 days ago
@RyanKidd Thanks for your comment.
You're right that I drew heavily from TARA's grant proposal. Doing that without citing TARA's grant application was unfair to their effort, and I should have acknowledged their work.
I’ve reached out to Yanni to offer a direct apology and ask if they’re comfortable with STARA acknowledging them explicitly in the proposal, and/or they prefer further changes to the name or proposal content.