·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Thousand Token Wood: shipping a multi-agent economy on a 3B model2h◆Startup Battlefield 200 applications officially close in 3 days5h◆Google will pay SpaceX $920M per month for compute6h◆The most interesting startups right now want to get you off your phone7h◆This is your laptop… on AI8h◆New York lawmakers pass one-year ban on new data centers9h◆The token bill comes due: Inside the industry scramble to manage AI’s runaway costs10h◆The latest AI news we announced in May 202610h◆The ‘together tech’ wave might be the most intriguing startup bet of 202611h◆This AI startup says it can tell if a script will make a hit film11h◆AirTrunk commits $30B to build 5GW of AI data centers in India12h◆The Meta hack shows there’s more to AI security than Mythos16h◆Mira Murati steps back into the spotlight, carefully20h◆SFMambaNet: Spectral-Frequency Enhanced Selective State Space Model for Correspondence Pruning21h◆Optical-Guided Neural Collapse for SAR Few-Shot Class Incremental Learning21h◆Dynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models21h◆Temporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents21h◆Why Muon Outperforms Adam: A Curvature Perspective21h◆Vision Hopfield Memory Networks21h◆Provably Auditable and Safe LLM Agents from Human-Authored Ontologies21h◆Thousand Token Wood: shipping a multi-agent economy on a 3B model2h◆Startup Battlefield 200 applications officially close in 3 days5h◆Google will pay SpaceX $920M per month for compute6h◆The most interesting startups right now want to get you off your phone7h◆This is your laptop… on AI8h◆New York lawmakers pass one-year ban on new data centers9h◆The token bill comes due: Inside the industry scramble to manage AI’s runaway costs10h◆The latest AI news we announced in May 202610h◆The ‘together tech’ wave might be the most intriguing startup bet of 202611h◆This AI startup says it can tell if a script will make a hit film11h◆AirTrunk commits $30B to build 5GW of AI data centers in India12h◆The Meta hack shows there’s more to AI security than Mythos16h◆Mira Murati steps back into the spotlight, carefully20h◆SFMambaNet: Spectral-Frequency Enhanced Selective State Space Model for Correspondence Pruning21h◆Optical-Guided Neural Collapse for SAR Few-Shot Class Incremental Learning21h◆Dynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models21h◆Temporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents21h◆Why Muon Outperforms Adam: A Curvature Perspective21h◆Vision Hopfield Memory Networks21h◆Provably Auditable and Safe LLM Agents from Human-Authored Ontologies21h◆
News/Internal Knowledge Without External Expression: Probing the Generalization Boundary of a Classical Chinese Language Model
arxiv
PublishedApril 17, 2026 at 4:00 AM
—neutral

Internal Knowledge Without External Expression: Probing the Generalization Boundary of a Classical Chinese Language Model

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2604.14180v1 Announce Type: new Abstract: We train a 318M-parameter Transformer language model from scratch on a curated corpus of 1.56 billion tokens of pure Classical Chinese, with zero English characters or Arabic numerals. Through systematic out-of-distribution (OOD) testing, we investigat

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

// no spam · unsubscribe one-click · free forever

Discussion
Mentioned models
01
  • 01
    Transformer
Source
↗
arxiv
Read original ↗All from arxiv →
Tags
03
#language-models#uncertainty#metacognition

No replies yet. Be first.

Mentioned models
01
  • 01
    Transformer
Source
↗
arxiv
Read original ↗All from arxiv →
Tags
03
#language-models#uncertainty#metacognition

Related coverage

More from ARXIV
arxivSFMambaNet: Spectral-Frequency Enhanced Selective State Space Model for Correspondence Pruning21harxivOptical-Guided Neural Collapse for SAR Few-Shot Class Incremental Learning21harxivDynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models21harxivTemporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents21h
The Bubble Brief
WEEKLY

Read language-models insights every Tuesday — top movers, new releases, story of the week.

// no spam · unsubscribe one-click · free forever

Originally published on arxiv ↗
HomeModelsNews