paper

llm-training ,  harness ,  paper ,  agent ,  llm ,  self-improvement ,  recursive-self-improvement ,  distributed-training

ai-ml-papers-of-the-week ,  paper ,  ai-security ,  automatic-prompt-optimization ,  jepa ,  serverbench ,  platonic-representation-hypothesis ,  talentrl ,  hierarchical-skill-routing ,  vibevideobench ,  sema-auto ,  subject-matter-expert-critique ,  hybrid-computer-use ,  cepo ,  astra-bench

image-captioning ,  multimodal ,  representation-learning ,  self-supervised ,  paper ,  scaling-law ,  jepa

ai-ml-papers-of-the-week ,  small-language-model ,  paper ,  looped-transformer ,  recursive-self-improvement ,  agentbaton ,  aw-0 ,  agentic-world-action-models ,  transferable-tool-selection ,  the-trinity ,  capability-aware-harness-router ,  attribution-gradients ,  interactiveworld ,  latent-numerical-frame ,  terminus-4b ,  protocols-for-agentic-ai

ai-ml-papers-of-the-week ,  paper ,  llm-reasoning ,  self-evolving-agents ,  discoveryrsi ,  videoabc ,  raw-rag ,  harness-scaling ,  zeroth-order-optimization ,  logra ,  oo-harness ,  video-complexity-measurement ,  diagnostic-analysis-for-llm-reasoning ,  hybrid-reasoning-agents

pytorch ,  paper ,  flow-matching ,  speculative-decoding ,  llm

paper ,  ai ,  llm ,  ai-agent

quantum-computing ,  reinforcement-learning ,  paper ,  quantum-error-correction ,  google ,  magic-state-cultivation ,  willow

multi-token-prediction ,  self-evolving-agent ,  ai-ml-papers-of-the-week ,  paper ,  multimodal-rag ,  decision-model ,  videojev ,  harness2 ,  gemini-4-argon ,  discoloop ,  pixsearch ,  google-earth-ai-gra ,  harmony-harness ,  longharness ,  vepa ,  personalized-llm-recommendation ,  autoregressive-beam-search ,  geospatial-reasoning-agent ,  agent-guardrails

fine-tuning ,  paper ,  sakana-ai ,  agent ,  llm ,  multi-agent ,  self-improvement ,  recursive-self-improvement

paper ,  ai-for-science ,  code-generation ,  world-model ,  agent ,  llm ,  self-improvement ,  jepa

latex ,  claude-code ,  claude-code-skill ,  paper ,  arxiv ,  translation ,  agent-skills

ai-ml-papers-of-the-week ,  paper ,  recursive-self-improvement ,  scholarevolve ,  raw-video-mid-training ,  lime ,  lime-xor ,  flowagent ,  sigma ,  veriharness ,  vhop-router ,  vhop ,  multi-step-embedding-retrieval ,  visual-rag ,  sclate ,  continual-learning-agent ,  agent-harness-optimization ,  llm-chat-template-as-voice-switch

fine-tuning ,  lora ,  calibration ,  paper ,  jev ,  llm ,  qwen

paper ,  vllm ,  diffusion ,  tts ,  inference-framework ,  vllm-omni ,  omni-modality ,  serving

paper ,  google-research ,  labor-market ,  ai-economics ,  generative-ai ,  google ,  ai-productivity ,  nber

inference ,  paper ,  looped-transformer ,  apple ,  llm ,  contrastive-decoding ,  decoding

tool-use ,  reinforcement-learning ,  paper ,  agent ,  mcp ,  benchmark ,  evaluation

paper ,  jev ,  reranking ,  llm ,  qwen ,  recommender-system ,  latency

ai-ml-papers-of-the-week ,  paper ,  jev ,  spatial-intelligence ,  recommendation-system ,  jev-mem ,  looped-transformers ,  agent-world ,  loopcd ,  unievo-vl ,  kilometervision ,  epistemic-depth ,  digital-consciousness-model ,  agentic-idea-management ,  automated-research ,  rlvr ,  rltl-dr ,  autogym ,  agent-gyms ,  skillgym

open-vocabulary ,  paper ,  computer-vision ,  dataset ,  benchmark ,  scene-graph-generation ,  visual-relationship-detection ,  dinov3

ai-ml-papers-of-the-week ,  autoresearch ,  paper ,  agent-harness ,  llm-wiki ,  personalized-llm ,  agentic-memory ,  agentic-reasoning ,  on-device ,  harness-zero ,  sol-pi ,  jev-mem ,  wiki-foundation-model ,  think-like-a-world-model-act-like-a-vla ,  regularized-recursive-self-improvement ,  slopshape ,  lige-gr ,  generative-recommendation ,  omni-modal-embeddings ,  ovis-embedding

retrieval ,  reinforcement-learning ,  paper ,  google ,  diffusion ,  icml ,  recommendation ,  generative-retrieval

ai-ml-papers-of-the-week ,  paper ,  personalized-llm ,  compile-by-training ,  harness-design-for-coding-agents ,  agora ,  shared-memory-for-autoresearch ,  rl-compiled-diffusion ,  thinking-with-looped-flows ,  verifiable-social-reasoning ,  fuse-framework-for-user-mediated-social-evaluation ,  procedural-graphs ,  self-evolving-execution-structures ,  nova-kv ,  end-of-token ,  breaking-the-token-ceiling ,  tailored-to-you

openai ,  paper ,  ai ,  vidraft ,  llm ,  비드래프트 ,  arxiv

technical-report ,  reinforcement-learning ,  paper ,  llm ,  transformer ,  architecture ,  state-tracking

ai-ml-papers-of-the-week ,  paper ,  prompt-optimization ,  declarative-attention ,  auto-recsys ,  agentgrad ,  embedders-dilemma ,  amber ,  e-commerce-bench ,  topoagent ,  echo-grpo ,  mind-viruses ,  sdd ,  videoecho-r1 ,  synthetic-data-for-time-series-foundation-models

self-evolving-agent ,  ai-ml-papers-of-the-week ,  paper ,  self-improving-agents ,  tpr-tensor-product-representations ,  discover ,  gradient-symbol-systems ,  meta-n ,  harness-of-harness ,  scroll ,  programmatic-context-management ,  harnessdev ,  qwen-drive ,  terminal-universe ,  skillzip ,  scalable-agent-skill-libraries ,  recevolve

zero-shot ,  forecasting ,  paper ,  foundation-model ,  google-research ,  google ,  time-series

agent ,  cybersecurity ,  alignment ,  technical-report ,  openai ,  paper ,  llm ,  safety