ข่าว
How S&P is using deep web scraping, ensemble learning and Snowflake architecture to collect 5X more data on SMEs
Previously, S&P only had data on about 2 million SMEs, but its AI-powered RiskGauge platform...
How runtime attacks turn profitable AI into budget black holes
AI inference attacks drain enterprise budgets, derail regulatory compliance and destroy new AI deployment ROI.Read...
How OpenAI’s red team made ChatGPT agent into an AI fortress
Discover OpenAI’s red team blueprint: How 110 coordinated attacks and 7 exploit fixes created ChatGPT...
How Much Can We Forget about Data Contamination?
arXiv:2410.03249v4 Announce Type: replace-cross Abstract: The leakage of benchmark data into the training data has...
How Do LLMs Really Reason? A Framework to Separate Logic from Knowledge
Unpacking Reasoning in Modern LLMs: Why Final Answers Aren’t Enough Recent advancements in reasoning-focused LLMs...
Hospital cyber attacks cost $600K/hour. Here’s how AI is changing the math
How Alberta Health Services is using advanced AI to bolster its defenses as attackers increasingly...
Hope Speech Detection in code-mixed Roman Urdu tweets: A Positive Turn in Natural Language Processing
arXiv:2506.21583v1 Announce Type: new Abstract: Hope is a positive emotional state involving the expectation of...
Highlighted at CVPR 2025: Google DeepMind’s ‘Motion Prompting’ Paper Unlocks Granular Video Control
Key Takeaways: Researchers from Google DeepMind, the University of Michigan & Brown university have developed...
High-Entropy Token Selection in Reinforcement Learning with Verifiable Rewards (RLVR) Improves Accuracy and Reduces Training Cost for LLMs
Large Language Models (LLMs) generate step-by-step responses known as Chain-of-Thoughts (CoTs), where each token contributes...
High Accuracy, Less Talk (HALT): Reliable LLMs through Capability-Aligned Finetuning
arXiv:2506.04051v1 Announce Type: new Abstract: Large Language Models (LLMs) currently respond to every prompt. However...
Handling Numeric Expressions in Automatic Speech Recognition
arXiv:2408.00004v2 Announce Type: replace-cross Abstract: This paper addresses the problem of correctly formatting numeric expressions...
GuidedBench: Measuring and Mitigating the Evaluation Discrepancies of In-the-wild LLM Jailbreak Methods
arXiv:2502.16903v2 Announce Type: replace Abstract: Despite the growing interest in jailbreak methods as an effective...
