LIVE PULSE
4.0 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.2 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.0 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src1.8 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.4 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.1 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.1 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.1 Study examines issue bias in LLMs used as writing assistants before Swedish 2026 election — 1 src1.1 Study Audits Misalignment in Multi-Modal World Models — 1 src1.1 Retrieval-Grounded Reasoning Approach Proposed for Universal Multimodal Embeddings — 1 src4.0 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.2 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.0 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src1.8 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.4 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.1 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.1 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.1 Study examines issue bias in LLMs used as writing assistants before Swedish 2026 election — 1 src1.1 Study Audits Misalignment in Multi-Modal World Models — 1 src1.1 Retrieval-Grounded Reasoning Approach Proposed for Universal Multimodal Embeddings — 1 src
HEATPULSEAI MAGAZINES
FLIP · FOLLOW · SAVE
Log inSign up

Video Generation

topic5 events
papersTODAY 04:00 UTC

Study finds video models store correct physics but fail to apply it

A new arXiv paper asks whether video generators that produce physically implausible motion never learned the correct dynamics or merely fail to use what they learned. The authors introduce a notion of causal writability to probe this, and report that the correct motion is still represented inside the model and can be made to steer generation. The finding points to a gap between internal knowledge and how it is used, rather than a simple absence of physical understanding.

video-models/physics/interpretability/generative-models
papersTODAY 04:00 UTC

arXiv paper proposes training paradigm for fast long video generation

A new arXiv preprint addresses the difficulty of generating coherent minute-long videos, noting that while short clips are plentiful and high quality, long-form training data is scarce and confined to a few domains. The authors propose a training approach that combines mode-seeking and mean-seeking objectives to speed up long video generation. The work is positioned as a way to overcome the data bottleneck that limits scaling from seconds to minutes.

video-generation/research/training/generative-models
papersTODAY 04:00 UTC

Method Turns Sequenced Fuzzy Cognitive Maps into Causal Virtual Worlds via Video Generators

A new arXiv paper describes an approach for building and steering causal virtual worlds using large language and video model agents. It relies on feedback fuzzy cognitive maps to capture the detailed causal structure of the simulated environment. The technique converts sequenced FCMs into worlds that video generators can render.

research/video-generation/agents/simulation
papersSEP 12 04:00 UTC

Benchmark and Method Proposed for Think-with-Video Reasoning in Generative Models

A new arXiv paper argues that while video generation models now produce convincing and temporally consistent output, it is unclear whether they can reason through video by following symbolic rules, obeying physics, and working toward defined goals. The authors introduce a benchmark for measuring this think-with-video ability and propose an approach for improving it. The work is listed under the cs.AI cross-submission category.

video-generation/benchmarks/reasoning/arxiv
papersSEP 10 04:00 UTC

AgenticGen: Reward-Guided Agentic Video Generation for Advertising

A new arXiv preprint introduces AgenticGen, a framework that treats advertising video creation as an agentic reasoning task shaped by reward signals rather than simple clip synthesis. The method conditions generation on a specific product and aims to optimize for online business outcomes, building on video foundation models that can produce realistic footage from multimodal inputs. The paper appears as a cross-listing in arXiv's AI and computational linguistics categories.

agents/video-generation/advertising/rewards