REVIEW 22 cited by
AdaCoT: Pareto-Optimal Adaptive Chain-of-Thought Triggering via Reinforcement Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
AdaCoT: Pareto-Optimal Adaptive Chain-of-Thought Triggering via Reinforcement Learning
read the original abstract
Large Language Models (LLMs) have demonstrated remarkable capabilities but often face challenges with tasks requiring sophisticated reasoning. While Chain-of-Thought (CoT) prompting significantly enhances reasoning, it indiscriminately generates lengthy reasoning steps for all queries, leading to substantial computational costs and inefficiency, especially for simpler inputs. To address this critical issue, we introduce AdaCoT (Adaptive Chain-of-Thought), a novel framework enabling LLMs to adaptively decide when to invoke CoT. AdaCoT framed adaptive reasoning as a Pareto optimization problem that seeks to balance model performance with the costs associated with CoT invocation (both frequency and computational overhead). We propose a reinforcement learning (RL) based method, specifically utilizing Proximal Policy Optimization (PPO), to dynamically control the CoT triggering decision boundary by adjusting penalty coefficients, thereby allowing the model to determine CoT necessity based on implicit query complexity. A key technical contribution is Selective Loss Masking (SLM), designed to counteract decision boundary collapse during multi-stage RL training, ensuring robust and stable adaptive triggering. Experimental results demonstrate that AdaCoT successfully navigates the Pareto frontier, achieving substantial reductions in CoT usage for queries not requiring elaborate reasoning. For instance, on our production traffic testset, AdaCoT reduced CoT triggering rates to as low as 3.18\% and decreased average response tokens by 69.06%, while maintaining high performance on complex tasks.
Forward citations
Cited by 22 Pith papers
-
Thinking Without Words: Efficient Latent Reasoning with Abstract Chain-of-Thought
Abstract-CoT lets models reason with short discrete latent token sequences from a reserved vocabulary, using warm-up training and RL to match verbal CoT performance with up to 11.6x fewer tokens.
-
Beyond Majority Voting: Towards Fine-grained and More Reliable Reward Signal for Test-Time Reinforcement Learning
SCOPE uses step-wise confidence and dynamic subgroups to create finer pseudo-labels in test-time RL, delivering 13.1% relative gains on AIME 2025 over majority-voting baselines.
-
APIVOT: Adaptive Planning with Interleaved Vision-Language Thoughts
A VLM planner that adaptively inserts latent visual thoughts of future states into its reasoning trace beats language-only and prior VLM planners on long-horizon kitchen tasks, especially under tight free space.
-
DyCon: Dynamic Reasoning Control via Evolving Difficulty Modeling
DyCon dynamically controls reasoning depth in LRMs by modeling evolving difficulty from step-level embeddings, reducing redundant steps across multiple benchmarks.
-
Efficient Agentic Reasoning Through Self-Regulated Simulative Planning
SR²AM achieves competitive Pass@1 accuracy on diverse tasks with 25.8-95.3% fewer reasoning tokens than much larger models by using self-regulated simulative planning trained via supervised learning and RL.
-
Taming the Thinker: Conditional Entropy Shaping for Adaptive LLM Reasoning
CES applies conditional bidirectional entropy control on top of DAPO to improve accuracy and shorten responses on mathematical benchmarks for 7B and 1.5B LLMs.
-
Case-Based Calibration of Adaptive Reasoning and Execution for LLM Tool Use
CAST extracts complexity and failure profiles from historical tool-use trajectories to drive adaptive reasoning and fine-grained rewards in RL, yielding up to 5.85 pp higher execution accuracy and 26% shorter reasonin...
-
CoWorld-VLA: Thinking in a Multi-Expert World Model for Autonomous Driving
CoWorld-VLA encodes world information into four expert tokens that condition a diffusion-based planner, yielding competitive collision avoidance and trajectory accuracy on the NAVSIM benchmark.
-
CoWorld-VLA: Thinking in a Multi-Expert World Model for Autonomous Driving
CoWorld-VLA extracts semantic, geometric, dynamic, and trajectory expert tokens from multi-source supervision and feeds them into a diffusion-based hierarchical planner, achieving competitive collision avoidance and t...
-
CODA: Difficulty-Aware Compute Allocation for Adaptive Reasoning
CODA uses rollout-based difficulty signals to drive two gates that penalize verbosity on easy instances and promote deliberation on hard ones, cutting token use over 60% on simple tasks while maintaining accuracy.
-
Inverse IFEval: Can LLMs Unlearn Stubborn Training Conventions to Follow Real Instructions?
A new 1,012-question benchmark shows LLMs often fail instructions that deliberately invert common training conventions, revealing a measurable gap in counterintuitive instruction following.
-
R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning
R-4B uses bi-mode annealing plus a GRPO variant with forced thinking and non-thinking rollouts so a 4B multimodal model adaptively picks when to reason, reportedly matching 16B thinking models on several reasoning benchmarks.
-
Structured Thoughts For Improved Reasoning And Context Pruning
Structured try/outcome SFT improves math reasoning by up to 8% over standard SFT and enables pruning ~85% of context with ~9% accuracy drop.
-
Spatial Reasoning via Modality Switching Between Language and Symbolic Representation
Introduces a trustworthiness-and-complexity switching metric that lets LLMs choose between language and grid modalities for spatial reasoning, yielding up to 42% gains in tested settings.
-
Spatial Reasoning via Modality Switching Between Language and Symbolic Representation
Introduces a modality-switching mechanism for LLMs on spatial reasoning tasks using a trustworthiness and complexity based metric, showing up to 42% performance improvement.
-
Learning When to Reason for Text-to-SQL via SFT and DPO
A training pipeline that teaches a text-to-SQL model to skip chain-of-thought reasoning on easy queries preserves accuracy while reducing inference tokens and latency.
-
ThinkDial: An Open Recipe for Controlling Reasoning Effort in Large Language Models
An open training recipe that teaches a 32B LLM to reason in three adjustable modes, cutting token use by half or three quarters with small accuracy losses.
-
Think in Blocks: Adaptive Reasoning from Direct Response to Deep Reasoning
A training framework that makes LLMs predict an integer reasoning budget as a number of blocks and adapt reasoning depth to task difficulty.
-
SuCo: Sufficiency-guided Continuous Adaptive Reasoning
SuCo defines minimal sufficient CoT and applies a two-stage fine-tuning plus RL framework to enable continuous adaptive reasoning control, claiming gains in both accuracy and token efficiency on math, code, and science tasks.
-
PlanE: Meta Planning of Data, Tuning, and Inference for Extractive-based LLMs
A quadratic meta-planner trained on a few model-dataset runs selects the optimal data-tuning-inference configuration for extractive LLMs, matching grid search on three IE tasks.
-
Seedream 4.0: Toward Next-generation Multimodal Image Generation
Seedream 4.0 unifies text-to-image synthesis, image editing, and multi-image composition in an efficient diffusion transformer pretrained on billions of pairs and accelerated to 1.8 seconds for 2K output.
-
Reinforcement Learning Meets Large Language Models: A Survey of Advancements and Applications Across the LLM Lifecycle
A survey that maps reinforcement learning methods, datasets, benchmarks, and open-source tools across the full training lifecycle of large language models, focusing on verifiable-reward reasoning.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.