·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Bringing Value Models Back: Generative Critics for Value Modeling in LLM Reinforcement Learning42m◆Subagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks42m◆Distribution-Consistent Inference for Dynamic Sparse Mixture-of-Experts42m◆In RAG We Trust? Measuring Robustness of Retrieval-Augmented Generation Under Document Poisoning42m◆FrontierChallenge: Evaluating Scientific Workflow Completion42m◆IBIB: A Protocol for Measuring Enterprise AI Systems by Serving Route, Not Model Identifier42m◆AgenticGen: Reward-Guided Agentic Video Generation for Advertising42m◆Strangers to Themselves: What Language Models Say About Themselves Is Generic42m◆Omni Interaction Agent Technical Report42m◆CoGReV: A Confidence-Gated Post-Hoc Non-Monotonic Belief Revision Framework for Phishing Website Classification42m◆Bit-Flip Attacks on Vision-Language-Action Models: Action-Decoding Architecture Shapes the Vulnerability42m◆Can Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven Policy Operationalization42m◆RiLM: Parameter-Efficient Language Modeling via Geodesic Decoding42m◆Tracing Computation Density in LLMs42m◆Cultural Binding Heads in Language Models42m◆Direct Diversity Optimization for Diverse Successful Trajectories in Preference Post-Training42m◆Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models42m◆Palmyra x6 Technical Report: An Agentic, Tool-Use Model Post-Trained via Anchored Supervised Fine-Tuning42m◆'Ghaib in Translation' aka Unseen Harm: Measuring Cross-Script Safety Inconsistency with 'Missed-in-Urdu' Scores in LLM Hate Speech Detection42m◆DexterSQL: Deep Schema Exploration and Rule-based Correction for Text-to-SQL Generation42m◆Bringing Value Models Back: Generative Critics for Value Modeling in LLM Reinforcement Learning42m◆Subagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks42m◆Distribution-Consistent Inference for Dynamic Sparse Mixture-of-Experts42m◆In RAG We Trust? Measuring Robustness of Retrieval-Augmented Generation Under Document Poisoning42m◆FrontierChallenge: Evaluating Scientific Workflow Completion42m◆IBIB: A Protocol for Measuring Enterprise AI Systems by Serving Route, Not Model Identifier42m◆AgenticGen: Reward-Guided Agentic Video Generation for Advertising42m◆Strangers to Themselves: What Language Models Say About Themselves Is Generic42m◆Omni Interaction Agent Technical Report42m◆CoGReV: A Confidence-Gated Post-Hoc Non-Monotonic Belief Revision Framework for Phishing Website Classification42m◆Bit-Flip Attacks on Vision-Language-Action Models: Action-Decoding Architecture Shapes the Vulnerability42m◆Can Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven Policy Operationalization42m◆RiLM: Parameter-Efficient Language Modeling via Geodesic Decoding42m◆Tracing Computation Density in LLMs42m◆Cultural Binding Heads in Language Models42m◆Direct Diversity Optimization for Diverse Successful Trajectories in Preference Post-Training42m◆Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models42m◆Palmyra x6 Technical Report: An Agentic, Tool-Use Model Post-Trained via Anchored Supervised Fine-Tuning42m◆'Ghaib in Translation' aka Unseen Harm: Measuring Cross-Script Safety Inconsistency with 'Missed-in-Urdu' Scores in LLM Hate Speech Detection42m◆DexterSQL: Deep Schema Exploration and Rule-based Correction for Text-to-SQL Generation42m◆
News/Decoder-Side Semantic Conditioning for Low-Bitrate Neural Speech Compression
arxiv
PublishedSeptember 10, 2026 at 4:00 AM

Decoder-Side Semantic Conditioning for Low-Bitrate Neural Speech Compression

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

arXiv:2512.21653v2 Announce Type: replace-cross Abstract: Speech codecs are usually optimized for waveform fidelity, allocating bits to acoustic detail that can be inferred from linguistic structure. This leads to inefficient compression and degraded recognition performance. We propose SemDAC, a sem

Stay posted· Newsletter

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

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

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
arxivBringing Value Models Back: Generative Critics for Value Modeling in LLM Reinforcement Learning42marxivSubagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks42marxivDistribution-Consistent Inference for Dynamic Sparse Mixture-of-Experts42marxivIn RAG We Trust? Measuring Robustness of Retrieval-Augmented Generation Under Document Poisoning42m
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

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

Originally published on arxiv ↗
HomeModelsNews