Tagged articles

early exit

3 articles · Page 1 of 1

Jul 16, 2026 · Artificial Intelligence

Brake Overthinking in Long‑Reasoning Models by Detecting Semantic Redundancy

Long‑thinking LLMs often waste 41‑52% of tokens after the final answer; the PUMA framework detects when reasoning stops producing new semantic information, enabling early exit that cuts average token usage by 26.2% while keeping accuracy stable and even improving speed across multiple benchmarks.

PUMAearly exitllm-inference

0 likes · 9 min read

Brake Overthinking in Long‑Reasoning Models by Detecting Semantic Redundancy

Tencent Tech

Oct 27, 2025 · Artificial Intelligence

How SpecExit Cuts Large Reasoning Model Inference Time by Up to 2.5×

SpecExit combines early‑exit and speculative decoding to let large reasoning models detect when they have almost finished thinking, trimming redundant chain‑of‑thought steps, reducing over‑thinking by 72% and achieving up to 2.5× faster end‑to‑end inference without noticeable accuracy loss.

AIInference AccelerationSpeculative Decoding

0 likes · 6 min read

How SpecExit Cuts Large Reasoning Model Inference Time by Up to 2.5×

Java Captain

Nov 25, 2018 · Fundamentals

Optimizing Bubble Sort with Early Termination Using a Boolean Flag

The article recounts a interview scenario where a candidate struggled with a basic bubble sort implementation, then explains the standard bubble sort algorithm, its inefficiencies, and demonstrates an optimization that adds a boolean flag to detect a sorted pass and terminate early, improving performance.

Javaalgorithm optimizationbubble sort

0 likes · 5 min read