읽을거리&정보공유

paper ,  llm ,  anthropic ,  safety ,  interpretability ,  alignment ,  emotion

ai-coding-tools ,  zed ,  geeknews ,  claude-code ,  openrouter ,  cursor

paper ,  rl-with-verifiable-rewards ,  ai-ml-papers-of-the-week ,  asi-evolve ,  skillclaw ,  ai-for-ai ,  agentic-evolver ,  functional-emotions ,  adaptive-token-allocation ,  neural-computers ,  tempo ,  autoregressive-models ,  energy-based-models ,  next-token-prediction ,  megatrain ,  ai-assistance-reduces-persistence ,  visionclaw ,  self-distilled-rlvr ,  rlsd-rlvr-with-self-distiilation ,  small-vlm

cuda ,  gpu ,  linear-attention ,  pytorch ,  flash-attention ,  ml-engineering ,  gated-delta-networks ,  gpu-kernel ,  gated-linear-attention ,  kimi-delta-attention ,  lightning-attention ,  cuda-linear-attention ,  cula

agent ,  claude ,  api ,  anthropic ,  advisor-tool ,  agentic-patterns ,  sub-agent

browser ,  chrome-extension ,  webgpu ,  on-device ,  gemma ,  inference ,  llm-on-browser

paper ,  agent ,  agentic-ai ,  multi-agent-system ,  bpm ,  agentic-business-process-management

ai-agent ,  orchestration ,  agent-teams ,  multi-agent ,  geeknews ,  claude-code ,  agent-harness ,  multi-agent-orchestration

lora ,  apple-silicon ,  peft ,  mps ,  fine-tuning ,  pytorch ,  multimodal ,  gemma

graphrag ,  rag ,  python ,  knowledge-graph ,  lightrag ,  hkuds ,  emnlp

code-agent ,  andrej-karpathy ,  knowledge-graph ,  graphify ,  knowledge-graph-skill ,  safi-shamsi

embedding ,  multimodal ,  contrastive-learning ,  qwen3 ,  bidirlm ,  gemma3 ,  weight-merging ,  omnimodal-encoders ,  omnimodal ,  huggingface

memory ,  local-ai ,  semantic-search ,  mcp ,  opensource ,  knowledge-graph ,  ai-memory ,  chromadb

litellm

text-to-speech ,  nvidia ,  speech-to-speech ,  moshi ,  personaplex ,  full-duplex ,  voice-ai

cybersecurity ,  claude ,  anthropic ,  product-launch

multi-agent ,  typescript ,  autonomous-agent ,  cli ,  inkos ,  novel-generation ,  creative-writing

typescript ,  ai-agent ,  context-engineering ,  meta-prompting ,  spec-driven ,  autonomous ,  gsd ,  get-shit-done ,  context-rot ,  gsd-build

fine-tuning ,  paper ,  google-research ,  recommendation ,  bayesian-inference ,  bayesian

local-ai ,  voice-assistant ,  on-device ,  gemma ,  parlor ,  multimodal-ai ,  kokoro-tts ,  gemma-4-e2b ,  voice-llm ,  fikrikarim

rag ,  vector-db ,  pageindex ,  retrieval ,  reasoning ,  information-retrieval

skills ,  arxiv ,  machine-learning ,  pytorch ,  claude-code ,  paper2code ,  research ,  claude-code-skill

legalize-kr ,  korean-law

git-worktree ,  multi-agent ,  typescript ,  coding-agents ,  parallel ,  yc-w26 ,  desktop-app ,  emdash ,  llm-provider-agnosticism ,  agentic-development-environment

reinforcement-learning ,  humanoid-robot ,  latent ,  motion-capture ,  sim-to-real ,  ppo

deep-learning ,  transformer ,  machine-learning ,  neural-network ,  primer ,  lecture

test-time-compute ,  paper ,  survey-paper ,  trustworthy-llm ,  latent-space ,  ai-ml-papers-of-the-week ,  recommendation-system ,  latent-generalization ,  zero-error-horizon ,  clawkeeper ,  abuse-detection-lifecycle ,  hyperalign ,  diffusion-model-alignment ,  gistbench ,  adaptive-generative-modeling ,  instance-specific-specialization ,  expectation-error-bounds-for-transfer-learning