Hugging Face Blog
Real-Time Intelligence with IBM Time Series Models on Confluent
Real-Time Intelligence with IBM Time Series Models on Confluent
// espacio-latente.com / radar
Píldora diaria de lo que se mueve en IA: laboratorios, papers, blogs de referencia y releases relevantes, resumido y con link al original. Se actualiza dos veces al día. · ← volver a espacio-latente.com · archivo · RSS
Real-Time Intelligence with IBM Time Series Models on Confluent
ATV Big Air Tour turned 3 days of work into 3 hours with ChatGPT
llm-gemini 0.34
Claude's new system prompt really doesn't want to reproduce song lyrics
Google releases Gemini 3.8 Flash, its third Flash model in six weeks
Trump may be forced to reveal secret rules feds use for AI safety testing
Facilitating AI integration with simplicity at scale
PRO-Step: Step-level Process Reward Optimization for Retrieval-Augmented Generation
Learning Evidence Sufficiency Boundaries for Selective Answering in Grounded Multi-Hop QA
SpeakPay: Domain-Adaptive LoRA Fine-Tuning of Whisper for Low-Resource Nepali Financial Speech Recognition
MemeCULT-1K: Benchmarking South Asian Cultural Context and Humor Understanding of Multimodal Models
VakyArth: Evaluating Pragmatic Competence in LLMs across Indic Languages
Disentangling Statistical Preemption from Entrenchment in Language Models' Avoidance of Overgeneralization
How Do Prompt Variations Affect Energy Consumption in On-Device LLMs?
TalkFa: A Unified Benchmark for Farsi Dialogue Generation and Understanding
AVERT: Audio-Verified Adjudication for Spoken Dialogue State Tracking
Interpretable Symptom Vectors for Depression in a Large Language Model
Candidate Generation and Definition-Guided Verification for Sentence-Level Depression Symptom Recognition
Cite or Decline: A Strict Course-Grounded Chatbot for STEM Lecture Videos
Thinking effort aligns between humans and reasoning models in abductive reasoning
GAPS: Dimension-Level Gates for Conditional Activation Steering
Grounded, Compute-Efficient LLM Policy Agents for Energy-Poverty Equity in Physically-Constrained Peer-to-Peer Energy Markets
Sparse Readout Prism: Explaining Logit-Lens Scores in Features Instead of Tokens
NS-Copilot: An LLM-Driven Agent System for Autonomous Neuroscience Analysis
How Output Format Confounds Data Quality and Capability in Instruction Tuning
A Tri-Agent Framework for Evaluating and Aligning Question Clarification Capabilities of Large Language Models
Qwen 3.8 27B available on Cerebras at 1500 tok/SEC
.name Termination
OpenAI begins rolling out GPT-6 Astra
K2 Horizon: A connected fleet of six open models
Any Human Ever – One life, drawn at random from all who have ever lived
Porting my 1993 Amiga game to Godot, with an LLM reading the 68000 assembly
Ask HN: Why were OpenAI, Claude, and Grok simultaneously down?
Static Allocation, Constant Work
Audacity 4.0
Gooseworks (YC W23) Is Hiring – Founding Creative Engineer
Google Antigravity TOS: 3rd party usage can get Google account suspended
Introducing WeatherNext 3, our most advanced and accurate global weather AI model
[AINews] Muse Spark 1.3 matches GPT-5.6-Sol, confirming Meta Superintelligence as the newest Frontier Lab, >90% discount for training
OpenAI’s next big AI model has ‘entered the AGI era’
Nvidia launches free tool that links idle computers into a personal AI data center
Google now lets you chat with Gmail, Docs, and Keep
ChatGPT, Grok, and Claude all went down at the same time
Google says its AI weather model is getting better
Nvidia is buying Hugging Face for almost $13 billion
Google says its new Gemini 3.8 Flash model ‘works harder’ but might cost more
Abliteration.ai is making a business out of removing AI guardrails
Meta is paying to peek at how you use their latest AI model
OpenAI launches Astra, its powerful (and controversial) new model
Ollie is betting its focus on privacy can help it win the AI assistant race
Google’s latest AI weather model gives you no excuse to forget your umbrella
Nvidia confirms it will buy Hugging Face for $12.9 billion
The Builders Stage brings practical strategies for scaling startups to TechCrunch Disrupt 2026
Palo Alto Networks paid $500M for Thrive-backed Console, sources say
TechCrunch Disrupt 2026’s new Real World AI Stage features Nvidia, robots, and extinct animals
OpenAI’s new reasoning technique alarms AI safety experts
WMLLM: Self-Evolving Optimization Agents via Predict-Then-Act World Modeling
DiDrive: A Risk-Aware Hierarchical Diffusion Framework for Safe Offline Reinforcement Learning in Autonomous Driving
Prompt-Space Meta-Learning Does Not Transfer Across Users: A Frozen-LLM Negative Result
Construir este día costó 55.955 tokens de entrada y 0 de salida en 59 llamadas a modelos — unos 0,0560 $.