Jeonhwi Kang
4 papers · 2026–2026 · 2 venues · across top CS/AI venues
Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.
Top co-authors
Research topics
Keywords
model merging
(1)
commonsense qa
(1)
reasoning length
(1)
large language model
(1)
reflective reasoning
(1)
wall-clock speedup
(1)
semi-autoregressive decoding
(1)
sparse intervention
(1)
gsm8k
(1)
neurons
(1)
difficulty-aware
(1)
active-block tokens
(1)
llada-8b-instruct
(1)
soft-future context
(1)
stability contrast
(1)
answer-forcing utility estimation
(1)
overthinking-mitigating rescaling
(1)
ruar
(1)
utility-guided advantage rescaling
(1)
activation contrast
(1)
Papers
DALSA: Difficulty-Aware Long-to-Short LLM Reasoning with Model Merging
AACL 2026
Probing Reasoning Length Control in Language Models via Neurons
AACL 2026
Signals from the Future: Future-aware Semi-Autoregressive Decoding for Diffusion Language Models
EMNLP 2026
When Reflection Helps: Utility-Guided Advantage Rescaling for Efficient Reasoning
EMNLP 2026