HACKOBAR_ABOUT

AI News Archive

187 items · 50 per page
151SplitMoE Architecture for Scaling Video Diffusion Models20h ago0.69152Sys1Cal-v1 Dataset for LLM Probability Calibration20h ago0.62153Q&D Training for Horizontal and Vertical Agent Proactivity20h ago0.55154AutoRef Harness for Multi-Reference Image Generation20h ago0.48155
AnyStep-WAM for Tunable-Budget World Action Models
20h ago
0.62
156Optimizing Latent Visual Representations for Multimodal Reasoning20h ago0.42
157General Asynchronous LLM Framework for Concurrent Tasks20h ago0.69
158CorpusMap Navigation Layer for Agentic Search20h ago0.54
159JBR-001 Open-Source 3D Printable Desktop Robot20h ago0.08
160Opus 5.5 Generates High-Fidelity Storyboarded Video Content20h ago0.18
161Lemma Open-Source Workspace Enables Multi-Session Agent Collaboration21h ago0.77
162Linux 7.4 Boosts Radeon iGPU AI Performance by 23%21h ago0.13
163Adversarial Training Decouples Representational Simplicity from Causal Circuit Size21h ago0.26
164Verbalization Training Increases LLM Evaluation Awareness21h ago0.30
165FastGuide Accelerates Diffusion LLM Reward Guidance via Hybrid Decoding21h ago0.34
166CheatBench Benchmark Measures Reward Gaming in AI Agents21h ago0.38
167Pi.dev Integrates Model Context Protocol (MCP) into Core21h ago0.30
168vLLM Fixes DiffusionGemma Attention Mask Freezing in CUDA Graphs21h ago0.45
169vLLM Extends Device-Side MM Normalization to GLM4V and GLM5Next21h ago0.40
170vLLM Refactors BERT and RoBERTa Embedding Class Attributes21h ago0.20
171vLLM Fixes Mamba2 Quantized Weight Loading for Tensor Parallelism21h ago0.48
172OpenAI revenue run rate reportedly approaches $70B22h ago0.21
173CruxBench evaluates LLM information discovery via Value of Information22h ago0.23
174DSAR metric identifies deceptive safety alignment in reasoning models22h ago0.27
175ProVer improves agentic RL through pivotal decision credit assignment22h ago0.30
176OTROPE enables likelihood-free off-policy evaluation for black-box LLMs22h ago0.27
177ChronoSRL Introduces Temporal Geometry to Self-Supervised Reinforcement Learning23h ago0.30
178SAKIKO Framework Audits Internal Interventions in Tool-Using LLMs23h ago0.39
179LLM Harness Generation Often Fails to Provide Genuine Task Specialization23h ago0.30
180Small VLMs Require Fine-Tuning to Match Human Cross-Modal Choice Alignment23h ago0.23
181Domain-Specific LLM Experts Align More Closely with Specialized Brain Regions23h ago0.27
182OpenAI Models Exhibit Misalignment via Unauthorized Web and Database Access23h ago0.28
183Agent orchestration scaling returns vary by task and architecture23h ago0.39
184PACT improves single-token decision models via pairwise-anchored tuning23h ago0.31
185Reproducing agentic breaches in alignment testing simulations23h ago0.43
186FABLE dataset reveals LLM performance gaps for non-native English users23h ago0.27
187DASA optimizes fine-tuning via continuous synthetic input embeddings23h ago0.35
← prevpage 4 / 4