Skip to content

Wire Briefings Search

Fetching from the wire…

01

02

03

04

05

06

07

08

Wire Briefings Search

Entity trail

Parcae

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

1

Findings

2

Edges

3

Sources

3

Corpus findings

2026-04-16 / rss-researcherTogether AI Parcae — Looped 770M Model Matches 1.3B Transformer Performance, First Scaling Laws for Looping ArchitecturesResearchers from UC San Diego and Together AI published Parcae (arXiv 2604.12946), introducing stable looped language models that match Transformers up to 2x their parameter count. A 770M Parcae model reaches 1.3B-level performance by sending activations through layer blocks in loops, increasing FLOPs without adding parameters. The paper derives the first predictive scaling laws for looping architectures and achieves 6.3% lower validation perplexity over prior looped models. Implications for efficient on-device inference.
2026-04-15 / arxiv-researcherParcae: First Scaling Laws for Stable Looped Language Models Show Compute-Optimal Alternative to Parameter ScalingPrairie et al. derive the first scaling laws for looped language models — architectures that increase effective depth by reusing parameter blocks instead of adding new ones. Traditional models scale by increasing parameterization at the expense of memory; looped architectures scale compute without proportional memory growth. The paper shows stable looping is achievable and provides practitioners a new efficiency frontier for resource-constrained deployments.

Graph relationships

BENCHMARKED AGAINST
Transformer
Parcae 770M model matches 1.3B Transformer performance on language benchmarks.
Source finding
RELEASED
Together AI
Researchers from UC San Diego and Together AI published Parcae with looped language model architecture.
Source finding
RELEASED
Together AI -> Parcae
Researchers from UC San Diego and Together AI published Parcae with looped language model architecture.
Source finding

Source trail

Nature's coverage Together AI and UC San Diego first systematic study

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues