·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
CompMat-Bench: Benchmarking AI Agents for Computational Materials Science6h◆LLMR: Knowledge Distillation with a Large Language Model-Induced Reward6h◆PRISM: PRIvileged Sibling-guided Multi-Agent Topology Design6h◆Training Reasoning Models to Generate Diverse Scientific Ideas6h◆OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling6h◆MagiC: Evaluating Multimodal Cognition Toward Grounded Visual Reasoning6h◆Causal Syntax Circuits in CodeLMs: Budget and Probe Decide What Looks Localised6h◆Learning Topology-Aware Representations for Mechanics-Relevant TPMS Metamaterials6h◆Selection-Safe Conformal Early Exit: Distribution-Free Coverage under Adaptive Depth6h◆Sliced Wasserstein Sketches with Error Certificates6h◆Generalization Guarantees of Randomized Block Coordinate Descent for Nonconvex Learning6h◆Query-Efficient Rejection Sampling for Parallel Masked Diffusion6h◆Explore Once, Evaluate Many: Multiple Policy Evaluation for Stochastic Linear Bandits6h◆BitSparse: Compact Activation Storage for ReLU Feed-Forward Networks6h◆Two Sample Sizes in Test-Time LLM Ensembling: Buffered ERM and Minimax Theory6h◆ManiFusion: Unlocking the High-Throughput Generation via Superposition in Manifold Space6h◆Policy-Aligned Predictive Verification for Adaptive VLA Action-Chunk Execution6h◆DuET: Note-Frame Dual Encoding for Piano Performance-to-Score Transcription6h◆Size-Transferability of Graph Transformers with Convolutional Positional Encodings6h◆Where to Steer: Input-Dependent Layer Selection for Steering Improves LLM Alignment6h◆CompMat-Bench: Benchmarking AI Agents for Computational Materials Science6h◆LLMR: Knowledge Distillation with a Large Language Model-Induced Reward6h◆PRISM: PRIvileged Sibling-guided Multi-Agent Topology Design6h◆Training Reasoning Models to Generate Diverse Scientific Ideas6h◆OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling6h◆MagiC: Evaluating Multimodal Cognition Toward Grounded Visual Reasoning6h◆Causal Syntax Circuits in CodeLMs: Budget and Probe Decide What Looks Localised6h◆Learning Topology-Aware Representations for Mechanics-Relevant TPMS Metamaterials6h◆Selection-Safe Conformal Early Exit: Distribution-Free Coverage under Adaptive Depth6h◆Sliced Wasserstein Sketches with Error Certificates6h◆Generalization Guarantees of Randomized Block Coordinate Descent for Nonconvex Learning6h◆Query-Efficient Rejection Sampling for Parallel Masked Diffusion6h◆Explore Once, Evaluate Many: Multiple Policy Evaluation for Stochastic Linear Bandits6h◆BitSparse: Compact Activation Storage for ReLU Feed-Forward Networks6h◆Two Sample Sizes in Test-Time LLM Ensembling: Buffered ERM and Minimax Theory6h◆ManiFusion: Unlocking the High-Throughput Generation via Superposition in Manifold Space6h◆Policy-Aligned Predictive Verification for Adaptive VLA Action-Chunk Execution6h◆DuET: Note-Frame Dual Encoding for Piano Performance-to-Score Transcription6h◆Size-Transferability of Graph Transformers with Convolutional Positional Encodings6h◆Where to Steer: Input-Dependent Layer Selection for Steering Improves LLM Alignment6h◆
News/Evidence First, Arithmetic Second: A System Report and Failure Analysis for DocSem
arxiv
PublishedOctober 2, 2026 at 4:00 AM

Evidence First, Arithmetic Second: A System Report and Failure Analysis for DocSem

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

arXiv:2609.39013v1 Announce Type: new Abstract: EVICALC, our system for the DocSem shared task, achieved 8.61% joint accuracy on 1,730 tasks in the official final test evaluation. It reads a PDF, selects a passage, asks a language model to write an arithmetic expression, and evaluates that expressio

Stay posted· Newsletter

Three short emails a week: Monday, Wednesday and Friday. Top movers, new models and the most-referenced story.

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

Your email is used only to send The Bubble Brief. Privacy policy

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivCompMat-Bench: Benchmarking AI Agents for Computational Materials Science6harxivLLMR: Knowledge Distillation with a Large Language Model-Induced Reward6harxivPRISM: PRIvileged Sibling-guided Multi-Agent Topology Design6harxivTraining Reasoning Models to Generate Diverse Scientific Ideas6h
The Bubble Brief
MON · WED · FRI

Read AI insights every Monday, Wednesday and Friday — top movers, new models, story of the week.

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

Your email is used only to send The Bubble Brief. Privacy policy

Originally published on arxiv ↗
PrivacyTermsRefundsLegal notice|AboutMethodologyContact

© 2026 Databubble

Built by Marouane Gazouzi · Readyo, AI interview coach

HomeModelsNews