https://deepbrief.co/ai-research/neuroscientist-free-will-brain-imaging DeepBrief en 2026-04-13T16:04:50.764Z Does Free Will Exist? A Neuroscientist's Decade-Long Search 2026-04-13T16:30:00.743Z https://deepbrief.co/ai-society/sam-altman-home-attacked-twice DeepBrief en 2026-04-13T09:50:02.411Z Sam Altman's Home Targeted in Second Attack Within 48 Hours 2026-04-13T15:19:59.013Z https://deepbrief.co/ai-research/ai-models-cards-against-humanity-study DeepBrief en 2026-04-13T04:52:10.955Z AI Models Play Cards Against Humanity — and Agree With Each Other More Than With Humans 2026-04-13T06:30:12.270Z https://deepbrief.co/ai-research/graph-parsers-outperform-llms-relation-extraction DeepBrief en 2026-04-13T04:52:09.017Z LLMs Lose Ground to Lightweight Graph Parsers When Relation Extraction Gets Complex 2026-04-13T06:30:11.671Z https://deepbrief.co/ai-research/preference-training-data-ai-reasoning DeepBrief en 2026-04-13T04:52:07.086Z New Research Reveals What Makes Preference Training Data Work for AI Reasoning 2026-04-13T06:30:11.115Z https://deepbrief.co/ai-research/ai-writing-tools-linguistic-fingerprints DeepBrief en 2026-04-13T04:50:31.213Z AI Writing Tools Erasing Linguistic Fingerprints From Research Papers, Study Finds 2026-04-13T06:30:09.384Z https://deepbrief.co/ai-research/ai-agents-multi-user-failures DeepBrief en 2026-04-13T04:50:29.195Z Study Exposes Systematic Failures When AI Agents Serve Multiple Users 2026-04-13T06:30:08.823Z https://deepbrief.co/ai-research/ai-models-sentiment-analysis-gaza DeepBrief en 2026-04-13T04:50:27.197Z AI Models Disagree Sharply on Sentiment in Gaza War Headlines, Study Finds 2026-04-13T06:30:08.253Z https://deepbrief.co/ai-research/tree-structured-sparsity-transformer-compute DeepBrief en 2026-04-13T04:50:25.193Z Tree-Structured Sparsity Cuts Transformer Compute to 5% Without Accuracy Loss 2026-04-13T06:30:07.711Z https://deepbrief.co/ai-research/attn-sampler-diffusion-language-models DeepBrief en 2026-04-13T04:50:23.183Z New 'Attn-Sampler' Algorithm Improves Decoding for Diffusion Language Models 2026-04-13T06:30:07.155Z https://deepbrief.co/ai-research/temperature-settings-reasoning-ai-models DeepBrief en 2026-04-13T04:50:21.175Z Temperature Settings Matter as Much as Prompting Strategy in Reasoning AI Models 2026-04-13T06:30:06.601Z https://deepbrief.co/ai-research/neural-models-text-to-speech-quality DeepBrief en 2026-04-13T04:50:19.135Z Neural Models Match Human Consistency in Text-to-Speech Quality Ratings 2026-04-13T06:30:06.035Z https://deepbrief.co/ai-research/bias-mitigation-ai-model-internals DeepBrief en 2026-04-13T04:50:17.118Z New Research Maps How Bias Mitigation Reshapes AI Model Internals 2026-04-13T06:30:05.464Z https://deepbrief.co/ai-research/ai-medical-reasoning-gap-llms DeepBrief en 2026-04-13T04:50:15.070Z Survey Exposes Gap Between AI Medical Exam Performance and Clinical Reasoning 2026-04-13T06:30:04.341Z https://deepbrief.co/ai-research/wand-framework-speech-synthesis-memory DeepBrief en 2026-04-13T04:50:13.049Z WAND Framework Cuts AI Speech Synthesis Memory Costs by 66% 2026-04-13T06:30:03.766Z https://deepbrief.co/ai-research/diffusion-language-models-safety-bypass DeepBrief en 2026-04-13T04:42:36.593Z Two-Step Attack Bypasses Safety Guardrails in Diffusion Language Models at 81% Success Rate 2026-04-13T06:30:03.198Z https://deepbrief.co/ai-research/ai-memory-limits-language-models DeepBrief en 2026-04-13T04:42:34.503Z New Research Maps Memory Limits in Simple AI Language Models 2026-04-13T06:30:02.571Z https://deepbrief.co/ai-research/syndocdis-synthetic-doctor-conversations-llm DeepBrief en 2026-04-13T04:42:32.410Z SynDocDis Framework Uses LLMs to Generate Synthetic Doctor-to-Doctor Conversations 2026-04-13T06:30:01.443Z https://deepbrief.co/ai-research/ai-text-language-diversity-framework DeepBrief en 2026-04-13T04:42:30.376Z Mathematical Framework Reveals How AI-Generated Text Is Reshaping the Public Record — and How to Stop It Degrading 2026-04-13T06:30:00.855Z https://deepbrief.co/ai-research/whisper-ai-hallucination-research DeepBrief en 2026-04-13T04:42:26.216Z New Theory Explains Why Whisper's Largest AI Models Hallucinate More Confidently 2026-04-13T06:00:12.591Z https://deepbrief.co/ai-research/alphalab-autonomous-ai-research-agent DeepBrief en 2026-04-13T04:42:24.162Z AlphaLab: Autonomous AI Research Agent Writes GPU Code Running 91x Faster Than PyTorch's Compiler 2026-04-13T06:00:12.038Z https://deepbrief.co/ai-research/ai-escalation-framework-llm-help DeepBrief en 2026-04-13T04:42:20.034Z When Should an AI Ask for Help? New Research Maps Escalation 2026-04-13T06:00:10.908Z https://deepbrief.co/ai-research/csattention-llm-inference-acceleration DeepBrief en 2026-04-13T04:40:43.683Z CSAttention Cuts LLM Inference Time by Up to 4.6x on Long Contexts 2026-04-13T06:00:09.786Z https://deepbrief.co/ai-research/ai-reasoning-failures-drto-training DeepBrief en 2026-04-13T04:40:41.723Z New Training Method Cuts AI Reasoning Failures From Slight Wording Changes 2026-04-13T06:00:06.984Z https://deepbrief.co/ai-research/quantum-classical-ai-molecule-generation DeepBrief en 2026-04-13T04:40:39.747Z Hybrid Quantum-Classical AI Achieves 100% Valid Molecule Generation 2026-04-13T06:00:06.417Z https://deepbrief.co/ai-research/ai-reasoning-benchmark-open-weight-failures DeepBrief en 2026-04-13T04:40:37.800Z New Benchmark Exposes Reasoning Failures in Open-Weight AI Models 2026-04-13T06:00:04.118Z https://deepbrief.co/ai-research/quanbench-llms-quantum-code-struggle DeepBrief en 2026-04-13T04:40:35.838Z QuanBench+: New Benchmark Reveals LLMs Still Struggle With Quantum Code 2026-04-13T06:00:03.529Z https://deepbrief.co/ai-research/ai-sign-language-recognition-breakthrough DeepBrief en 2026-04-13T04:40:33.885Z New AI Model Cracks Sign Language Recognition at Scale by Mimicking Linguistic Structure 2026-04-13T05:30:12.729Z https://deepbrief.co/ai-research/ai-uncertainty-multimodal-framework DeepBrief en 2026-04-13T04:40:31.939Z New AI Framework Teaches Models to Say 'Probably' Across Text, Audio, and Video 2026-04-13T05:30:06.161Z https://deepbrief.co/ai-research/insedit-video-ai-editing-training DeepBrief en 2026-04-13T04:40:29.988Z InsEdit Teaches Video AI to Follow Editing Instructions With Less Training Data 2026-04-13T05:30:05.033Z https://deepbrief.co/ai-research/ai-hallucinations-3d-robotic-agents DeepBrief en 2026-04-13T04:40:28.058Z New Technique Cuts AI Hallucinations in 3D Robotic Agents 2026-04-13T05:30:04.457Z https://deepbrief.co/ai-research/ai-art-evaluator-semantic-meaning DeepBrief en 2026-04-13T04:40:26.126Z New AI Evaluator Judges Generative Art on Meaning, Not Just Appearance 2026-04-13T05:30:03.894Z https://deepbrief.co/ai-research/wilddet3d-object-detection-categories DeepBrief en 2026-04-13T04:35:55.176Z WildDet3D Identifies Over 13,500 Object Categories in 3D From Single Photo 2026-04-13T05:30:03.310Z https://deepbrief.co/ai-research/ai-memory-drug-discovery-success DeepBrief en 2026-04-13T04:31:04.315Z New AI Memory System Boosts Drug Discovery Success Rate by 36% 2026-04-13T05:00:12.650Z https://deepbrief.co/ai-research/sage-benchmark-ai-execution-gap-2 DeepBrief en 2026-04-13T04:31:02.343Z SAGE Benchmark Exposes Gap Between AI Intent Recognition and Service Execution 2026-04-13T05:00:12.083Z https://deepbrief.co/ai-research/ai-agents-benchmark-web-calculation DeepBrief en 2026-04-13T04:31:00.326Z New Benchmark Exposes AI Agents' Blind Spot: Combining Web Research With Calculation 2026-04-13T05:00:11.534Z https://deepbrief.co/ai-research/camera-artist-ai-video-cinematography DeepBrief en 2026-04-13T04:30:58.294Z Camera Artist: New AI Framework Teaches Video Generation to Think Like a Cinematographer 2026-04-13T05:00:10.971Z https://deepbrief.co/ai-research/cognitive-pressure-physics-intuition-planning DeepBrief en 2026-04-13T04:30:56.268Z Humans Shift Physics Intuition and Planning Strategy Under Cognitive Pressure, Study Finds 2026-04-13T05:00:10.408Z https://deepbrief.co/ai-research/advantage-guided-diffusion-reinforcement-learning DeepBrief en 2026-04-13T04:30:54.253Z Advantage-Guided Diffusion Cuts Compounding Errors in Reinforcement Learning World Models 2026-04-13T05:00:09.849Z https://deepbrief.co/ai-research/hypergraph-neural-networks-logical-contradictions DeepBrief en 2026-04-13T04:30:52.303Z Hypergraph Neural Networks Could Slash the Cost of Finding Logical Contradictions in Complex Systems 2026-04-13T05:00:09.279Z https://deepbrief.co/ai-research/ai-agent-benchmark-inefficiencies DeepBrief en 2026-04-13T04:30:50.226Z New Benchmark Exposes Hidden Inefficiencies in AI Agents 2026-04-13T05:00:08.705Z https://deepbrief.co/ai-research/pilotbench-ai-flight-safety-benchmark DeepBrief en 2026-04-13T04:30:48.201Z PilotBench Reveals Trade-Off Between AI Accuracy and Controllability in Flight Safety Testing 2026-04-13T05:00:08.133Z https://deepbrief.co/ai-research/petite-ai-tutor-student-framework DeepBrief en 2026-04-13T04:30:46.250Z Tutor-Student AI Framework Matches Top Multi-Agent Systems at Lower Cost 2026-04-13T05:00:07.568Z https://deepbrief.co/ai-research/starpo-ai-logical-consistency-reasoning DeepBrief en 2026-04-13T04:30:35.141Z StaRPO Targets Logical Consistency in AI Reasoning, Not Just Correct Answers 2026-04-13T05:00:06.999Z https://deepbrief.co/ai-research/sppo-algorithm-reduces-ai-training-costs DeepBrief en 2026-04-13T04:30:33.187Z SPPO: New Training Algorithm Cuts Compute Costs for Teaching AI to Reason 2026-04-13T05:00:06.425Z https://deepbrief.co/ai-research/ai-reads-physics-equations-images DeepBrief en 2026-04-13T04:30:31.184Z New AI System Reads Physics Equations Directly From Scientific Images 2026-04-13T05:00:05.848Z https://deepbrief.co/ai-research/ai-agents-external-memory-research DeepBrief en 2026-04-13T04:30:29.132Z Researchers Prove AI Agents Can Use Environment as External Memory 2026-04-13T05:00:05.295Z https://deepbrief.co/ai-research/ai-planning-domain-feedback-search DeepBrief en 2026-04-13T04:30:27.123Z Researchers Use Feedback-Driven Search to Improve AI Planning Domain Generation 2026-04-13T05:00:04.714Z https://deepbrief.co/ai-research/courcelle-theorem-graph-logic-decision-diagrams DeepBrief en 2026-04-13T04:30:25.212Z Researchers Extend Courcelle's Theorem to Show Graph Logic Can Be Compactly Represented in Decision Diagrams 2026-04-13T05:00:04.153Z https://deepbrief.co/ai-research/ramp-reinforcement-learning-classical-planning DeepBrief en 2026-04-13T04:30:23.308Z RAMP System Blends Reinforcement Learning and Classical Planning 2026-04-13T05:00:03.580Z https://deepbrief.co/ai-research/ai-marketing-agents-autonomous-performance DeepBrief en 2026-04-13T04:30:21.414Z AI Marketing Agents Sustain Performance Without Oversight, 11-Month Study Finds 2026-04-13T05:00:03.000Z https://deepbrief.co/ai-research/ai-accuracy-enterprise-decision-framework DeepBrief en 2026-04-13T04:30:19.529Z New Framework Addresses AI Accuracy Problem in Enterprise Decision-Making 2026-04-13T05:00:01.821Z https://deepbrief.co/ai-research/openkedge-protocol-ai-agent-safety DeepBrief en 2026-04-13T04:30:17.592Z OpenKedge Protocol Proposes Cryptographic Safety for Autonomous AI Agents 2026-04-13T05:00:01.042Z