Jaewon Sok

According to our database, Jaewon Sok authored at least 5 papers in 2026.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of five.

2026

Hallucination as Commitment Failure: Larger LLMs Misfire Despite Knowing the Answer.

Je Won Yeom, Jaewon Sok, Heejun Kim, Seonghyeon Park, Jeongjae Park, Taesup Kim

CoRR, May, 2026

From Noise to Diversity: Random Embedding Injection in LLM Reasoning.

Heejun Kim, Seungpil Lee, Je Won Yeom, Jaewon Sok, Seonghyeon Park, Jeongjae Park, Taesup Kim, Sundong Kim

CoRR, May, 2026

Efficient Epistemic Uncertainty Estimation for Large Language Models via Knowledge Distillation.

Seonghyeon Park, Je Won Yeom, Jaewon Sok, Jeongjae Park, Heejun Kim, Taesup Kim

CoRR, February, 2026

Garbage Attention in Large Language Models: BOS Sink Heads and Sink-aware Pruning.

Jaewon Sok, Je Won Yeom, Seonghyeon Park, Jeongjae Park, Taesup Kim

CoRR, January, 2026

EpiCaR: Knowing What You Don't Know Matters for Better Reasoning in LLMs.

Je Won Yeom, Jaewon Sok, Seonghyeon Park, Jeongjae Park, Taesup Kim

CoRR, January, 2026