Daily AI news, sourcedUpdates / Research
Updates Research · Unconfirmed · 138 items · newest issue first · page 7 of 7
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- JudgeMoE aggregates LLM-judge score distributions instead of scalar scoresUnconfirmedarXiv cs.CLOct 7Research
- EMODE uses dynamic para-semantic experts for emotion-aware speech language modelsUnconfirmedarXiv cs.CLOct 7Research
- Paper stabilizes continual learning in language models with condition-anchored distillationUnconfirmedarXiv cs.CLOct 7Research
- Paper compares tree navigation without LLM summaries for long-document QAUnconfirmedarXiv cs.CLOct 7Research
- Paper asks what makes a latent structure in language models actionableUnconfirmedarXiv cs.CLOct 7Research
- Paper finds behavioral history beats personal descriptions for LLM synthetic personasUnconfirmedarXiv cs.AIOct 7Research
- Paper guides autonomous-driving LLM decisions with retrieval of Chinese philosophical principlesUnconfirmedarXiv cs.AIOct 7Research
- Paper audits black-box LLM agents using a surrogate model's log-probabilitiesUnconfirmedarXiv cs.AIOct 7Research
- Astra and Claude prove the best-known packing for 11 squares is optimal, per a Reddit postUnconfirmedr/singularityOct 7Research
- Reddit post says low-paid field research jobs may become AI-proofUnconfirmedr/singularityOct 7Research
- Reddit thread asks whether people feel bad for mathematicians
- Reddit post jokes about vibemathed.com after OpenAI's math manuscript release
- Reddit post shares a meme on mathematicians after AI's math advances
- Anthropic details Claude text watermark research
- Anthropic details Claude-discovered novel enzyme system
- Functional Gradient Descent with Adaptive Representations paper discussedUnconfirmedreddit-machinelearningSep 30Research
- ClashRoyaleAI: open-source deterministic Clash Royale simulator for RL
- Tauon optimizer claimed to outperform Muon on GPT-Mini