Blog
270개 중 1-12번째 포스트
![[논문 리뷰] LT2: Linear-Time Looped Transformers](/assets/images/blog/20260603-paper-2605-20670-lt2-linear-time-looped-transfo.jpg)
[논문 리뷰] LT2: Linear-Time Looped Transformers
Looped Transformers (LT) have emerged as a powerful architecture by iterating their layers multiple times before decoding the final token. However, pairing them with full attention retains quadratic c...
![[논문 리뷰] Hallucinations Undermine Trust; Metacognition is a Way Forward](/assets/images/blog/20260603-paper-2605-01428-hallucinations-undermine-trust.jpg)
[논문 리뷰] Hallucinations Undermine Trust; Metacognition is a Way Forward
Despite significant strides in factual reliability, errors -- often termed hallucinations -- remain a major concern for generative AI, especially as LLMs are increasingly expected to be helpful in mor...
![[논문 리뷰] AI Must Embrace Specialization via Superhuman Adaptable Intelligence](/assets/images/blog/20260603-paper-2602-23643-ai-must-embrace-specialization.jpg)
[논문 리뷰] AI Must Embrace Specialization via Superhuman Adaptable Intelligence
Everyone from AI executives and researchers to doomsayers, politicians, and activists is talking about Artificial General Intelligence (AGI). Yet, they often don't seem to agree on its exact definitio...
![[논문 리뷰] AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation](/assets/images/blog/20260602-paper-2605-28655-autoscientists-self-organizing.jpg)
[논문 리뷰] AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation
Scientific research proceeds through iterative cycles of hypothesis generation, experiment design, execution, and revision. AI agents can automate parts of this process, but existing approaches typica...
![[논문 리뷰] AI for Auto-Research: Roadmap & User Guide](/assets/images/blog/20260602-paper-2605-18661-ai-for-auto-research-roadmap-a.jpg)
[논문 리뷰] AI for Auto-Research: Roadmap & User Guide
AI-assisted research is crossing a threshold: fully automated systems can now generate research papers for as little as $15, while long-horizon agents can execute experiments, draft manuscripts, and s...
![[논문 리뷰] Neural Weight Norm = Kolmogorov Complexity](/assets/images/blog/20260602-paper-2605-10878-neural-weight-norm-kolmogorov-.jpg)
[논문 리뷰] Neural Weight Norm = Kolmogorov Complexity
Why does weight decay work? We prove that, in any fixed-precision regime, the smallest weight norm of a looped neural network outputting a binary string equals the Kolmogorov complexity of that string...
![[논문 리뷰] Language Models Need Sleep](/assets/images/blog/20260526-paper-2605-26099-language-models-need-sleep.jpg)
[논문 리뷰] Language Models Need Sleep
Transformer-based large language models are increasingly used for long-horizon tasks; however, their attention mechanism scales poorly with context length. To handle this, we study a sleep-like consol...
![[논문 리뷰] SkillOpt: Executive Strategy for Self-Evolving Agent Skills](/assets/images/blog/20260526-paper-2605-23904-skillopt-executive-strategy-fo.jpg)
[논문 리뷰] SkillOpt: Executive Strategy for Self-Evolving Agent Skills
Agent skills today are hand-crafted, generated one-shot, or evolved through loosely controlled self-revision, none of which behaves like a deep-learning optimizer for the skill, and none of which reli...
![[논문 리뷰] Tokenisation via Convex Relaxations](/assets/images/blog/20260524-paper-2605-22821-tokenisation-via-convex-relaxa.jpg)
[논문 리뷰] Tokenisation via Convex Relaxations
Tokenisation is an integral part of the current NLP pipeline. Current tokenisation algorithms such as BPE and Unigram are greedy algorithms -- they make locally optimal decisions without considering t...
![[논문 리뷰] Advancing Mathematics Research with AI-Driven Formal Proof Search](/assets/images/blog/20260524-paper-2605-22763-advancing-mathematics-research.jpg)
[논문 리뷰] Advancing Mathematics Research with AI-Driven Formal Proof Search
Large language models (LLMs) increasingly excel at mathematical reasoning, but their unreliability limits their utility in mathematics research. A mitigation is using LLMs to generate formal proofs in...
![[논문 리뷰] Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning](/assets/images/blog/20260524-paper-2605-21488-equilibrium-reasoners-learning.jpg)
[논문 리뷰] Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning
Scaling test-time compute by iteratively updating a latent state has emerged as a powerful paradigm for reasoning. Yet the internal mechanisms that enable these iterative models to generalize beyond m...
![[논문 리뷰] Hallucinations Undermine Trust; Metacognition is a Way Forward](/assets/images/blog/20260524-paper-2605-01428-hallucinations-undermine-trust.jpg)
[논문 리뷰] Hallucinations Undermine Trust; Metacognition is a Way Forward
Despite significant strides in factual reliability, errors -- often termed hallucinations -- remain a major concern for generative AI, especially as LLMs are increasingly expected to be helpful in mor...