Product
Help
Pricing
About
Jobs
Log in
Get started
Author
Add to collection
Daniel Tan
Observed name:
Tan, Daniel
Institutions:
Arcadia,
University College London
Observed institutions:
Guru Ghasidas Vishwavidyalaya,
Arcadia,
University College London
OpenAlex ID:
A5134864218
Conditional misalignment: common interventions can hide emergent misalignment behind contextual triggers
2026
·
Dubiński, Jan
, Betley, Jan
, et al.
·
arXiv (Cornell University)
·
18
PDF
Inoculation Adapters: Improved Selective Generalization of Capabilities with Fewer Surprising Backdoors
2026
·
Riché, Maxime
, Tan, Daniel
, et al.
·
arXiv (Cornell University)
·
4
PDF
Don't Inoculate Everything: Stratified Inoculation Prompting Narrows Backdoor Triggers and Preserves Desired Traits
2026
·
Dymkiewicz, Kajetan
, Farrelly, Tim
, et al.
·
arXiv (Cornell University)
·
1
PDF
Stress-testing Alignment Midtraining
2026
·
Baines, Sid
, Bostock, Jonathan
, et al.
·
arXiv (Cornell University)
·
1
PDF
Character Training for Risk-Averse Agents
2026
·
Dhoot, Arav
, Pandey, Punya Syon
, et al.
·
arXiv (Cornell University)
PDF
AI in Financial Planning - Opportunities and Challenges
2025
·
Tan, Daniel
·
Journal of International DBA Studies - GGU
PDF
Show more works
View as search
Works count:
7
Citation count:
25
H-index:
2
I10-index:
1
Year
Topic
Type
