RESEARCH

–

12 May 2026 15h20

TMAS Scaling Test-Time Compute via Multi-Agent Synergy

Qwen-Image-2.0 Technical Report

By

–

12 May 2026 15h14

Qwen-Image-2.0 Technical Report

→ View original post on X — @montreal_ai

AGI ALPHA: Scalable Substrate for Intelligence Organizations

By

@montreal_ai

–

12 May 2026 14h53

"AGI ALPHA: A Scalable Substrate for Intelligence Organizations" Vincent Boucher, President of http://
MONTREAL.AI and http://
QUEBEC.AI Paper: https://
github.com/MontrealAI/agi
alpha-first-real-loop/blob/main/docs/manuscript/AGI_ALPHA_Unified_Publication_Final.pdf
… #AGIAlpha #QuebecAI #SovereignAI

→ View original post on X — @jiqizhixin

Scal3R: New AI Research for Neural 3D Scene Reconstruction from Video

By

@jiqizhixin

–

12 May 2026 14h42

What if AI could reconstruct kilometer-scale 3D scenes from video as effortlessly as humans perceive a room? Researchers from Zhejiang University and Horizon Robotics present Scal3R. They introduce a neural context memory that compresses long-range scene info, adapting on the

→ View original post on X — @elliotchen100

The current proliferation of AI benchmarks

By

@elliotchen100

–

12 May 2026 13h49

Good grief, we really don't lack benchmarks

Soohak: A Benchmark for Evaluating Research-level Math in LLMs

By

–

12 May 2026 11h38

Soohak A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs

Sharing Papers and Apps on Hugging Face

By

–

12 May 2026 10h45

Paper:
https://huggingface.co/papers/2605.10922
…
App:
https://huggingface.co/spaces/TencentARC/Pixal3D
…

→ View original post on X — @jiqizhixin

PromptEcho: A New Zero-Annotation Reward Method for Text-to-Image Models

By

@jiqizhixin

–

12 May 2026 8h34

You could train text-to-image models without expensive human feedback! Alibaba Group and Zhejiang University researchers present PromptEcho—a reward method that uses a frozen vision-language model to measure image-prompt alignment directly, with zero annotations or extra

→ View original post on X — @latentspacepod

Thinking Machines launches native interaction model for realtime voice

By

@latentspacepod

–

12 May 2026 6h35

[AINews] TML-Interaction-Small 276B-A12Bhttps://t.co/Z8FHf217oU

Thinking Machines' Native Interaction Models – TML-Interaction-Small 276B-A12B – advances SOTA Realtime Voice and kills standard VAD

all our highlights! https://t.co/QagGmERPPw
— Latent.Space (@latentspacepod) 12 mai 2026

[AINews] TML-Interaction-Small 276B-A12B https://
latent.space/p/ainews-think
ing-machines-native-interaction
… Thinking Machines' Native Interaction Models – TML-Interaction-Small 276B-A12B – advances SOTA Realtime Voice and kills standard VAD all our highlights!

→ View original post on X — @emollick

True Superintelligence: Infer or Bypass Tacit Knowledge?

By

@emollick

–

12 May 2026 6h06

A true superintelligence should be able to infer or elicit tacit/contextual knowledge? Or route around it? Otherwise AI installations are still held back by Hayek, Coase, Weber, etc.: human constraints navigated by humans. Mediated adoption, not the "all jobs transformed" pitch