·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Ensemble forecasting with IBM Granite time-series models23m◆How to Use Parallel Data for Multilingual Pretraining4h◆A KAN-based Non-linear Factorization for Embedding Layer Compression4h◆How Much Gradient Does Truncated Backprop Lose in Looped Transformers?4h◆Orbitol: Crystal Generation on Wyckoff Orbits4h◆Differentially Private Sparse Reward Estimation with Preference Feedback4h◆Group Distributionally Robust Optimization over Decentralized Networks4h◆In-Context Bias Propagation in LLM-Based Tabular Data Generation4h◆Operon: Incremental Construction of Ragged Data via Named Dimensions4h◆A Generalisation Signal Need Not Be a Model-Selection Signal4h◆Mitigating Reward Hacking in Realistic Scenarios Through Representation Engineering4h◆Characterizing Oblivious Common-Reward Multi-Player Bandits: Universal Stationarity, Coordination Traps, and Certified Global Rates4h◆DriveMLP: When Training Strategy Outperforms Model Complexity in Autonomous Driving4h◆Hierarchical Exponential-Gaussian Mixtures for Watch-Time Distribution Prediction4h◆RL-IF: Tracking Non-local Training Data Influence for Online Reinforcement Learning4h◆A Model-Free Continuous-Time RL Algorithm from Discrete-Time Off-Policy Data with End-to-End Guarantees4h◆Where Do Legal Factors Live in LLMs?4h◆Transformers as Generic Deep Neural Networks: Exact Compilation, the Cost of Removing Activation-Activation Products, and Multi-Agent Attention Clustering4h◆S^3martCirc: Self-supervised Smart Circuit Discovery4h◆Unit-Aware Differentially Private Quantum Machine Learning for Clinical and Wearable Health Data4h◆Ensemble forecasting with IBM Granite time-series models23m◆How to Use Parallel Data for Multilingual Pretraining4h◆A KAN-based Non-linear Factorization for Embedding Layer Compression4h◆How Much Gradient Does Truncated Backprop Lose in Looped Transformers?4h◆Orbitol: Crystal Generation on Wyckoff Orbits4h◆Differentially Private Sparse Reward Estimation with Preference Feedback4h◆Group Distributionally Robust Optimization over Decentralized Networks4h◆In-Context Bias Propagation in LLM-Based Tabular Data Generation4h◆Operon: Incremental Construction of Ragged Data via Named Dimensions4h◆A Generalisation Signal Need Not Be a Model-Selection Signal4h◆Mitigating Reward Hacking in Realistic Scenarios Through Representation Engineering4h◆Characterizing Oblivious Common-Reward Multi-Player Bandits: Universal Stationarity, Coordination Traps, and Certified Global Rates4h◆DriveMLP: When Training Strategy Outperforms Model Complexity in Autonomous Driving4h◆Hierarchical Exponential-Gaussian Mixtures for Watch-Time Distribution Prediction4h◆RL-IF: Tracking Non-local Training Data Influence for Online Reinforcement Learning4h◆A Model-Free Continuous-Time RL Algorithm from Discrete-Time Off-Policy Data with End-to-End Guarantees4h◆Where Do Legal Factors Live in LLMs?4h◆Transformers as Generic Deep Neural Networks: Exact Compilation, the Cost of Removing Activation-Activation Products, and Multi-Agent Attention Clustering4h◆S^3martCirc: Self-supervised Smart Circuit Discovery4h◆Unit-Aware Differentially Private Quantum Machine Learning for Clinical and Wearable Health Data4h◆
News/One Policy, Many Styles: Where to Aggregate Preferences for Steerable Humanoid Control?
arxiv
PublishedOctober 22, 2026 at 4:00 AM

One Policy, Many Styles: Where to Aggregate Preferences for Steerable Humanoid Control?

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

arXiv:2610.28925v1 Announce Type: cross Abstract: Modern robotic agents are increasingly expected to support steerable control, where a single policy can adjust its behavioral style at runtime under changing user preferences. Preference-conditioned reinforcement learning provides this interface, but

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
arxivHow to Use Parallel Data for Multilingual Pretraining4harxivA KAN-based Non-linear Factorization for Embedding Layer Compression4harxivHow Much Gradient Does Truncated Backprop Lose in Looped Transformers?4harxivOrbitol: Crystal Generation on Wyckoff Orbits4h
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