REVIEW 14 cited by
Efficiently Scaling LLM Reasoning with Certaindex
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
read the original abstract
Test-time reasoning algorithms such as chain-of-thought, self-consistency, and MCTS enhance LLM problem-solving but can wastefully generate many tokens without improving accuracy. At the same time, we observe that these algorithms exhibit answer stabilization: their intermediate solutions often cease to change after a certain point, and further investment of compute does not change their final answer. To quantify this phenomenon, we introduce Certaindex, an algorithm-agnostic metric measuring this evolving stability, signaling when further computation is unlikely to alter the final result. Certaindex is lightweight, can accelerate reasoning program inference via early exit, and further enables dynamic token allocation, gang scheduling, and many opportunities when integrated with real-world LLM serving systems. To quantify real-world benefits, we built Certaindex as a scheduler into Dynasor, our reasoning-aware LLM serving system, and demonstrate up to 50% compute savings and 3.3x higher throughput in real workloads with no accuracy drop. Our code is available at https://github.com/hao-ai-lab/Dynasor.git
Forward citations
Cited by 14 Pith papers
-
SPORK: Self-Speculative Forking to Accelerate Agentic LLM Inference
A training-free self-fork of an agentic LLM predicts its own next tool call early enough to hide 16–37% tool-wait latency, cutting Qwen3-32B GAIA P95 by 18% with accuracy preserved.
-
BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokens
A control-token insertion and two-stage training method that lets LLMs adhere to user-specified reasoning token budgets while preserving math accuracy.
-
KVFlow: Efficient Prefix Caching for Accelerating LLM-Based Multi-Agent Workflows
KVFlow uses workflow-aware eviction priorities and overlapped KV prefetching to cut cache-miss latency in LLM multi-agent serving.
-
SmartThinker: Learning to Compress and Preserve Reasoning by Step-Level Length Control
SmartThinker shortens AI reasoning chains by estimating the importance of each step and penalizing the unimportant ones, preserving accuracy while using far fewer tokens.
-
Do Thinking Tokens Help or Trap? Towards More Efficient Large Reasoning Model
Suppressing 'thinking tokens' in a 1.5B reasoning model preserves accuracy while cutting tokens, and the proposed DuP-PO RL method improves both accuracy and efficiency over GRPO.
-
How Far Are We from Optimal Reasoning Efficiency?
The authors define a reasoning efficiency frontier and a gap metric (REG), then train models with REO-RL to shrink the gap by at least 50% with only small accuracy losses.
-
Kinetics: Rethinking Test-Time Scaling Laws
A memory-aware test-time scaling law shows small models are overestimated and sparse attention is needed for efficient scaling.
-
System-1.5 Reasoning: Traversal in Language and Latent Spaces with Dynamic Shortcuts
System-1.5 Reasoning lets LLMs reason in latent space with early exits and step-skipping, matching chain-of-thought accuracy at over 20x speedup on GSM8K and StrategyQA.
-
TrimR: Verifier-based Training-Free Thinking Compression for Efficient Test-Time Scaling
TrimR uses a lightweight verifier to detect redundant thoughts in reasoning models and truncates them, cutting runtime by up to 70% without retraining.
-
Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs
The survey's L1/L2 taxonomy and benchmark show that current reasoning models waste compute on easy problems and underthink hard ones, motivating more adaptive inference.
-
Don't Think Longer, Think Wisely: Optimizing Thinking Dynamics for Large Reasoning Models
DTO constructs compressed, ground-truth-curated reasoning trajectories from LRM outputs and uses them as preferred responses in SimPO, yielding up to 12% accuracy gains and roughly 40% token reduction on math benchmarks.
-
PATS: Process-Level Adaptive Thinking Mode Switching
PATS adapts the number of beam-search candidates per reasoning step using process reward model scores, improving accuracy-efficiency tradeoffs on math benchmarks.
-
PREMISE: Scalable and Strategic Prompt Optimization for Efficient Mathematical Reasoning in Large Models
The paper reports that a concise optimized prompt reduces reasoning tokens by up to 87.5% while roughly preserving accuracy on GSM8K, SVAMP, and MATH-500.
-
Towards Concise and Adaptive Thinking in Large Reasoning Models: A Survey
A comprehensive review that categorizes methods for shortening and adaptively triggering chain-of-thought reasoning in large language models.
Discussion (0). Sign in to comment.