BLADE: Boundary-Expanded and Layer-Adaptive Dynamic Exit for Efficient LLM Reasoning

• URL: http://arxiv.org/abs/2607.28966v1
• Date: Fri, 31 Jul 2026 02:36:55 GMT
• Title: BLADE: Boundary-Expanded and Layer-Adaptive Dynamic Exit for Efficient LLM Reasoning
• Abstract summary: BLADE terminates reasoning by estimating whether the generated prefix is sufficient for correct answering.<n> BLADE constructs multi-granular checkpoints from sentence, self-doubt, and paragraph boundaries.<n>Experiments show that BLADE preserves near-baseline accuracy while reducing generated tokens by 24.8% on Qwen3-8B and 15.8% on Qwen3-4B.
• Score: 12.482949673236563
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models often improve task performance by generating long reasoning traces, but the resulting computation is frequently wasted on redundant verification and revision. Existing probe-based early-exit approaches mainly inspect explicit self-doubt expressions, leaving many earlier termination opportunities undetected. Expanding inspection to ordinary reasoning boundaries improves coverage, but also exposes highly diverse intermediate states whose predictive information may reside in different hidden layers. We present Boundary-Expanded and Layer-Adaptive Dynamic Exit for Efficient LLM Reasoning (BLADE), a lightweight framework that dynamically terminates reasoning by estimating whether the generated prefix is sufficient for correct answering. BLADE constructs multi-granular checkpoints from sentence, self-doubt, and paragraph boundaries, and derives robust training labels through repeated answer completions. It further learns a compact subset of informative probe layers instead of relying on fixed choices or expensive representations from all layers. At inference time, calibrated predictions are combined with checkpoint-specific confirmation rules to balance responsiveness and premature-exit risk. Experiments on five benchmarks and two Qwen3 reasoning models show that BLADE preserves near-baseline accuracy while reducing generated tokens by 24.8% on Qwen3-8B and 15.8% on Qwen3-4B. Ablation studies further confirm the benefits of diverse checkpoints and automatic layer selection, demonstrating an effective approach to more efficient LLM reasoning.

Related papers

• GradCuit: Credit-Assigned Gradient Flow Enables Robust and Interpretable Test-Time Latent Reasoning [61.67411833252235]
  We introduce GradCuit, which inserts optimizable latent states at a selected Transformer layer between the hidden representations of the prompt and the generated continuation.<n>Across five instruction-tuned backbones, three reasoning benchmarks, and two answer formats, GradCuit achieves an average accuracy of 64.5%.<n>For interpretability, token-level gradient attribution reveals that latent influence concentrates on reasoning-connector tokens, while layer analysis identifies early-to-middle Transformer layers as the most effective optimization space.
  arXiv  Detail & Related papers  (2026-08-03T17:55:24Z)
• Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer Decoding [92.1521161575198]
  We introduce Confident Decoding, a training-free decoding strategy that dynamically selects the most reliable near-final layer.<n> Experiments across dense and Mixture-of-Experts LLMs demonstrate consistent gains on challenging reasoning benchmarks.
  arXiv  Detail & Related papers  (2026-06-20T07:03:26Z)
• What to Test Next: Interpretable Coverage Gap Discovery in Driving VLMs [52.50210189669399]
  Driving vision-language models (VLMs) must accurately understand scenes across diverse conditions defined by Operational Design Domains (ODDs)<n>We propose SliceScorer, a deterministic scoring rule for missing-slice recommendation.<n>We show that SliceNav surfaces high-risk coverage gaps more effectively than prior slice-discovery methods.
  arXiv  Detail & Related papers  (2026-06-01T03:18:01Z)
• Latent Abstraction for Retrieval-Augmented Generation [4.6096940605642915]
  Existing systems rely on generating natural language queries at each hop.<n>We propose textbfLAnR (Latent Abstraction for RAG), a unified framework in which a single LLM jointly performs encoding, retrieval, and generation entirely within its own latent space.
  arXiv  Detail & Related papers  (2026-04-20T06:26:13Z)
• Adaptive Conformal Prediction for Improving Factuality of Generations by Large Language Models [86.8650252164764]
  Large language models (LLMs) are prone to generating factually incorrect outputs.<n>We propose an adaptive conformal prediction approach that extends conformal score transformation methods to LLMs.<n>This enables prompt-dependent calibration, retaining marginal coverage guarantees while improving conditional coverage.
  arXiv  Detail & Related papers  (2026-04-15T15:35:42Z)
• Fine-Grained Uncertainty Quantification for Long-Form Language Model Outputs: A Comparative Study [0.0]
  We introduce a taxonomy for fine-grained uncertainty quantification in long-form LLM outputs.<n>We formalize several families of consistency-based black-box scorers, providing generalizations and extensions of existing methods.<n>We find 1) claim-response entailment consistently performs better or on par with more complex claim-level scorers, 2) claim-level scoring generally yields better results than sentence-level scoring, and 3) uncertainty-aware decoding is highly effective for improving the factuality of long-form outputs.
  arXiv  Detail & Related papers  (2026-02-19T15:02:29Z)
• Do LLMs Know They Are Being Tested? Evaluation Awareness and Incentive-Sensitive Failures in GPT-OSS-20B [1.948261185683419]
  We investigate whether "evaluation scent" inflates measured performance without commensurate capability gains.<n>We run six paired A/B scenarios that hold task content and decoding fixed while varying framing.<n>We provide a reproducible A/B framework (prompt banks, validators, per-run scores, scripts) and practical guidance.
  arXiv  Detail & Related papers  (2025-10-08T09:49:05Z)
• Eigen-1: Adaptive Multi-Agent Refinement with Monitor-Based RAG for Scientific Reasoning [53.45095336430027]
  We develop a unified framework that combines implicit retrieval and structured collaboration.<n>On Humanity's Last Exam (HLE) Bio/Chem Gold, our framework achieves 48.3% accuracy.<n>Results on SuperGPQA and TRQA confirm robustness across domains.
  arXiv  Detail & Related papers  (2025-09-25T14:05:55Z)
• Beyond Prompting: An Efficient Embedding Framework for Open-Domain Question Answering [22.121850498642008]
  Large language models have recently pushed open domain question answering to new frontiers.<n> prevailing retriever-reader pipelines often depend on multiple rounds of prompt level instructions.<n>We propose EmbQA, an embedding-level framework that enhances both the retriever and the reader.
  arXiv  Detail & Related papers  (2025-03-03T14:41:35Z)
• Self-Evaluation Improves Selective Generation in Large Language Models [54.003992911447696]
  We reformulate open-ended generation tasks into token-level prediction tasks. We instruct an LLM to self-evaluate its answers. We benchmark a range of scoring methods based on self-evaluation.
  arXiv  Detail & Related papers  (2023-12-14T19:09:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.