LLM News

Every LLM release, update, and milestone.

0
researchByteDance

Bytedance-led team argues three training steps are enough for AI agents, releases competitive open model

A team from Bytedance and several US universities argues in a new paper that three training steps are enough to build capable AI agents. The authors say stacked, multi-phase pipelines make it hard to tell which step helps and make reproduction expensive. According to The Decoder, the work is backed by a competitive open model.

0
benchmarkOpenAI

GPT-6.2 Astra scores 41.2% on XDOF's Ember-0 robot benchmark, up from 19.6% for GPT-6 Astra

Robotics infrastructure company XDOF reports that GPT-6.2 Astra reached a 41.2% average success rate across the 16 tasks of its Ember-0 benchmark, versus 19.6% for GPT-6 Astra. It is the first model to complete rope knot-tying and T-shirt unfolding on the benchmark. Average prediction latency fell from 15 seconds to 10.1 seconds.

3 min readvia qbitai.com ↗
0
research

MiniMax reports its unreleased MIM model solved 94 TII McEliece challenge instances, 55 previously unsolved

MiniMax published research on November 20 describing cryptanalysis results from MIM, an unreleased internal system. The company says MIM solved 94 TII McEliece challenge instances, including 55 it describes as previously unsolved, and states the work does not affect deployed cryptographic standards.

2 min readvia pandaily.com ↗
0
product updateReplit

Replit launches Tunnels to connect workspaces to private network databases and internal APIs

Replit has introduced Tunnels, a feature that connects Replit environments to private-network services such as databases and internal APIs through a client run by IT inside the network. Traffic is encrypted and authenticated with mTLS, and Tunnels integrate with Replit's connectors. Tunnels are in early access.

3 min readvia replit.com ↗
0
model release

Thinking Machines Lab releases Inkling 2 Mini: 26B-parameter looped MoE with 3B active and 1M-token context

Thinking Machines Lab has released Inkling 2 Mini, a multimodal Mixture-of-Experts model with about 26B physical total parameters and 3B active per token. It repeats its middle layers three times, supports a 1M-token context, and ships with a DSpark drafter. Hugging Face reports it leads 32 of 48 benchmarks against four comparable models.

Page 1 of 60Next →