Efficient Reasoning via Constrained Optimization in Latent Space
cs.AI, math.OC
Submitted: 2026-09-28
Updated: 2026-10-08
Code: https://github.com/hzn18/Opt4Reasoning
Terminology
Sources
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- When More is Less: Understanding Chain-of-Thought Length in LLMs
- Dynamic Early Exit in Reasoning Models
- AdaCtrl: Towards Adaptive and Controllable Reasoning via Difficulty-Aware Budgeting
- Not All Tokens Are What You Need In Thinking
- Long-Short Chain-of-Thought Mixture Supervised Fine-Tuning Eliciting Efficient Reasoning in Large Language Models
- AutoL2S: Auto Long-Short Reasoning for Efficient Large Language Models
- Think Only When You Need with Large Hybrid-Reasoning Models
- Reasoning Models Know When They're Right: Probing Hidden States for Self-Verification
- Internalizing LLM Reasoning via Discovery and Replay of Latent Actions
- Mitigating Overthinking in Large Reasoning Models via Manifold Steering
- ATLAS: Verifier-Guided Adaptive Latent Activation Steering for Efficient LLM Reasoning
- Training Verifiers to Solve Math Word Problems
- EasySteer: A Unified Framework for High-Performance and Extensible LLM Steering
- Continuous Chain of Thought Enables Parallel Exploration and Reasoning
- Soft Thinking: Unlocking the Reasoning Potential of LLMs in Continuous Concept Space
- CoT Vectors: Transferring and Probing the Reasoning Mechanisms of LLMs
- Compressed Chain of Thought: Efficient Reasoning Through Dense Representations
- Reasoning with Sampling: Your Base Model is Smarter Than You Think
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems