Search results for Neurips

Latest updates for Neurips

Fresh curated links around NeurIPS are collected here so marketers can spot useful updates and turn timely ideas into posts faster.

Post angles to try

Share the most useful takeaway for your audience.

Turn one article into a quick practical checklist.

Ask your audience how this shift affects their work.

Turn angles into scheduled posts

Fresh articles and ideas

Recent curated links from global sources. Generate one free draft from any story, then use SocialBu to schedule and refine your content calendar.

marktechpost.com /1 week ago

Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight...

Nous Research releases Contrastive Neuron Attribution (CNA), a method that identifies and ablates sparse MLP neuron circuits to steer LLM behavior — no sparse autoencoder training,...

Read source

salesforce.com /1 month ago

Salesforce AI Research at ICLR 2026

Salesforce AI Research will present 21 accepted papers at ICLR 2026, the Fourteenth International Conference on Learning Representations. The conference runs April 23–27 at the Rio...

Read source

marktechpost.com /1 week ago

NVIDIA AI Releases Nemotron-Labs-Diffusion: A Tri-Mode Language Model with 6× Tokens Per Forward Over Qwen3-8B

NVIDIA researchers have released Nemotron-Labs-Diffusion, a language model family that unifies three decoding modes in one architecture. The model supports autoregressive (AR) deco...

Read source

towardsdatascience.com /3 weeks ago

CSPNet Paper Walkthrough: Just Better, No Tradeoffs

A review of the Cross-Stage Partial Network paperâ€ŠÂ â€” â€Šand a from-scratch PyTorch implementation The post CSPNet Paper Walkthrough: Just Better, No Tradeoffs appeared first o...

Read source

marktechpost.com /3 weeks ago

NVIDIA AI Releases Star Elastic: One Checkpoint that Contains 30B, 23B, and 12B Reasoning Models with Zero-Shot Slicing

NVIDIA researchers have introduced Star Elastic, a post-training method that embeds multiple nested reasoning models вЂ” at 30B, 23B, and 12B parameter scales вЂ” inside a single c...

Read source

marktechpost.com /6 days ago

NVIDIA AI Releases Gated DeltaNet-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule

Linear attention squeezes the unbounded KV cache into a fixed-size recurrent state, but editing that memory without scrambling existing associations is hard. Prior delta-rule model...

Read source

marktechpost.com /3 weeks ago

Zyphra Introduces Tensor and Sequence Parallelism (TSP): A Hardware-Aware Training and Inference Strategy That Delivers...

Zyphra Introduces Tensor and Sequence Parallelism (TSP): A Folded Parallelism Strategy That Reduces Both Parameter and Activation Memory Across the Same GPU Axis The post Zyphra In...

Read source

marktechpost.com /3 days ago

NVIDIA Releases Polar, a Token-Faithful Rollout Framework for GRPO Training Across Codex, Claude Code, and Qwen Code

NVIDIA researchers have introduced Polar, a rollout framework that trains language agents using reinforcement learning without modifying their agent harnesses. Polar places a model...

Read source

marktechpost.com /3 days ago

Sakana AI Proposes DiffusionBlocks: a Block-wise Training Framework That Converts Residual Networks into Independently T...

DiffusionBlocks converts residual networks into independently trainable blocks by interpreting layer updates as reverse diffusion denoising steps. The post Sakana AI Proposes Diffu...

Read source

medium.com /1 week ago

Intersection of Neuroscience, AI/ML and economics

Continue reading on Medium »

Read source

marktechpost.com /4 weeks ago

A New NVIDIA Research Shows Speculative Decoding in NeMo RL Achieves 1.8× Rollout Generation Speedup at 8B and Projects...

A new paper from NVIDIA Research integrates speculative decoding directly into NeMo RL with a vLLM backend, delivering lossless rollout acceleration at both 8B and projected 235B m...

Read source

pandaily.com /2 weeks ago

ACL 2026: Alibaba DAMO Academy's I2B-LPO Breaks RLVR Homogenization — From Repetitive Sampling to Effective Exploration

Alibaba DAMO Academy's I2B-LPO framework, accepted at ACL 2026 Main, improves math reasoning accuracy by up to 5.3% and semantic diversity by 7.4% by guiding models to generate mor...

Read source

marktechpost.com /2 weeks ago

Nous Research Proposes Lighthouse Attention: A Training-Only Selection-Based Hierarchical Attention That Delivers 1.4–1....

Nous Research has published Lighthouse Attention, a selection-based hierarchical attention mechanism that wraps around standard scaled dot-product attention during pretraining and...

Read source

medium.com /3 weeks ago

Hopper: The Optimizer That Learns Parallelism 2x Faster Than Adam

Intro: Speeding Up IntelligenceContinue reading on Medium В»

Read source

roboticsandautomationnews.com /3 days ago

CVPR 2026 fields 16,000+ paper submissions on technical advances in AI

The program committee of the 2026 Conference on Computer Vision and Pattern Recognition (CVPR), one of the world’s leading artificial intelligence (AI) and computer vision research...

Read source

developer-tech.com /1 month ago

NVIDIA Nemotron 3 Nano Omni: Unifying multimodal AI inference

The launch of NVIDIA Nemotron 3 Nano Omni forces engineering teams to rethink multimodal AI deployment to maximise inference capacity. Agentic systems routinely process screen inte...

Read source

marktechpost.com /2 weeks ago

Nous Research Releases Token Superposition Training to Speed Up LLM Pre-Training by Up to 2.5x Across 270M to 10B Parame...

Nous Research releases Token Superposition Training (TST), a two-phase pre-training method that cuts wall-clock training time by up to 2.5x at matched FLOPs by averaging contiguous...

Read source

scmp.com /2 weeks ago

China deepens footprint at AI conference despite NeurIPS dispute, US tensions

Chinese technology companies and researchers turned out in force at a leading global artificial intelligence conference, despite mounting questions over whether they might avoid th...

Read source

marktechpost.com /2 weeks ago

Sakana AI and NVIDIA Introduce TwELL with CUDA Kernels for 20.5% Inference and 21.9% Training Speedup in LLMs

Sakana AI and NVIDIA Researchers demonstrate that simple L1 regularization can induce over 99% sparsity in feedforward layers with negligible downstream performance impact, and tra...

Read source

dev.to /1 month ago

Connecting Generative Adversarial Networks and Actor-Critic Methods

Read source

habr.com /2 weeks ago

Нейросуфлер

Всех нас очаровывают возможности ИИ описывать происходящее перед видеокамерой, особенно часто встречаются презентации Gemini. Но пока мы нигде не нашли ответа к вопросу – А зачем?...

Read source

habr.com /2 weeks ago

Нейросуфлер

Read source

habr.com /2 weeks ago

Нейросуфлер

Read source

marktechpost.com /1 week ago

NVIDIA Introduces a 4-Bit Pretraining Methodology Using NVFP4, Validated on a 12B Hybrid Mamba-Transformer at 10T Token...

NVIDIA introduces a 4-bit pretraining methodology built around the NVFP4 microscaling format — combining selective BF16 layers, 16×16 Random Hadamard Transforms on Wgrad inputs, 2D...

Read source

Turn fresh research into a full content calendar

Use SocialBu to discover ideas, generate post drafts, and schedule them across your social channels.