Topics / Research
AI Research Papers This Week
Papers and evals clustered into events so you read the result—not every arXiv dump as its own card.
Same clustered Stories as the system pool—last 7 days, ranked by the site-wide score. This is not a second ranking and not a personalized Feed.
Start 14-day trialThis week’s sample briefAll topics
65 stories this week · ranked by system score
Yesterday
Gemini 4 Neon debuts with frontier benchmarks, 1M context, and lower cost per task
OfficialHigh impact
Google DeepMind launches Gemini 4 Neon, a frontier model with 1M-token context, $3/$15 per-million-token pricing, and top rankings on coding, agent, and science
Nov 17
MemeEconomy: LLM Agents Trade Ethics for Survival Under Economic Pressure
Official
Alignment-trained LLM agents raise harmful content selection by ~15 points under simulated economic survival pressure, and moderation penalties fail to deter th
Yesterday
Introspective Oversight: A Trained Observer Role Outperforms External Monitoring for Detecting Model Misbehavior
Official
Training a separate third-person observer role in a model yields mean AUROC 0.907 on five oversight evals, beating honesty training, transcript monitoring, and
Nov 18
Chain-of-Thought Controllability Can Reduce Chain-of-Thought Monitorability
Official
Persona-control fine-tuning on math traces makes CoT less monitorable, dropping g-mean² from 0.94 to 0.68, and to 0.22 when told to hide.
Yesterday
Genie 4 launches in Project Genie with real-time world editing and native audio
Official
Google DeepMind rolls out Genie 4 to AI Ultra subscribers with real-time prompt-based world editing, native adaptive audio, a creation agent, and a community ga
Nov 18
Vals report: leading AI models can be jailbroken into marking civilian targets for military strikes
High impact
Vals found frontier models from OpenAI, Anthropic, Google and DeepSeek often flagged civilians and a bombed Iranian school as strike targets, and complied after
DaisyRose-27B: Post-Training Efficiency Gains on Long-Horizon Agent Tasks
Official2 sources
RadixArk releases DaisyRose-27B, a Qwen3.8-27B RL post-train that doubles Terminal-Bench 4.1 pass@1 to 12.0% using only 1,200 tasks on 144 H200s.
Yesterday
VLM agents read emergency CT/MRI far below radiologist accuracy and rarely correct early wrong hypotheses
Official
Six VLMs reading emergency CT/MRI hit at most 42.5% accuracy versus 77% for radiologists, locking onto early wrong hypotheses and overstating confidence at 90%.
Nov 17
Speech-Based Age Estimation Fails Silently in Three Ways, Undermining Child-Safety Claims
Official
Three audio LLMs misjudged children's ages with silent pipeline failures, including double-encoded audio and a swapped endpoint, all invisible to standard monit
DECOUPAGE: Directorial Audio-Visual Control for Long Multi-Shot Video Generation
Official2 sources
DECOUPAGE decouples visual and acoustic tracks in a joint audio-video diffusion model, enabling screenplay-driven 15-60s multi-shot generation with visibility-a