minbeomkim's picture

minbeomkim

mbkim

·

https://minbeomkim.github.io/

AI & ML interests

Safe, Reliable, and Controllable AI Agency.

Recent Activity

authored a paper 13 days ago

Beyond Normalization: Rethinking the Partition Function as a Difficulty Scheduler for RLVR

submitted a paper 20 days ago

CausalArmor: Efficient Indirect Prompt Injection Guardrails via Causal Attribution

authored a paper 21 days ago

LifeTox: Unveiling Implicit Toxicity in Life Advice

View all activity

Organizations

upvoted a paper 21 days ago

CausalArmor: Efficient Indirect Prompt Injection Guardrails via Causal Attribution

Paper • 2602.07918 • Published 24 days ago • 1

upvoted 2 papers 5 months ago

Drift: Decoding-time Personalized Alignments with Implicit User Preferences

Paper • 2502.14289 • Published Feb 20, 2025 • 1

FlowRL: Matching Reward Distributions for LLM Reasoning

Paper • 2509.15207 • Published Sep 18, 2025 • 117

upvoted a paper 6 months ago

Critic-Guided Decoding for Controlled Text Generation

Paper • 2212.10938 • Published Dec 21, 2022 • 2