Tagged articles

6 articles

Page 1 of 1

Apr 22, 2026 · Artificial Intelligence

Can LLMs Boost Reasoning Alone? Introducing SePT’s Simple Online Self‑Training

SePT (Self‑evolving Post‑Training) shows that a large language model can improve its mathematical reasoning ability by about ten percentage points using a reward‑free online self‑training loop that decouples generation temperature from standard SFT, matching or surpassing RL‑based methods without harming general performance.

LLMMathematical ReasoningOnline Learning

0 likes · 9 min read

Can LLMs Boost Reasoning Alone? Introducing SePT’s Simple Online Self‑Training

AI Insight Log

Mar 18, 2026 · Artificial Intelligence

MiniMax M2.7 Self‑Trains and Rivals GPT‑5 & Opus 4.6 on Eight Benchmarks

MiniMax M2.7, released just a month after M2.5, introduces a self‑evolution training loop and achieves competitive scores on eight benchmarks—matching or surpassing Claude Opus 4.6, GPT‑5.4, Sonnet 4.6 and Gemini 3.1 Pro—while showcasing autonomous skill building, multi‑agent collaboration, and real‑world productivity applications.

Agent TeamsClaude OpusGPT-5

0 likes · 10 min read

MiniMax M2.7 Self‑Trains and Rivals GPT‑5 & Opus 4.6 on Eight Benchmarks

Machine Learning Algorithms & Natural Language Processing

Mar 17, 2026 · Artificial Intelligence

80 Million Records Expose AI‑Generated Data Pollution Undermining Diagnostic Reliability

A large‑scale study of over 800,000 synthetic clinical records shows that self‑training loops of AI‑generated medical text, reports, and images cause severe loss of pathological diversity, vocabulary, and diagnostic confidence, prompting the authors to propose mixed‑real‑data training and quality‑aware filtering as mitigations.

Diagnostic reliabilityMitigationSelf‑Training

0 likes · 10 min read

80 Million Records Expose AI‑Generated Data Pollution Undermining Diagnostic Reliability

Tencent Advertising Technology

Feb 5, 2026 · Artificial Intelligence

How Multi-Agent VLMs and PNU Loss Achieve High‑Accuracy Harmful Content Detection with Only 50 Labels

This article presents a low‑resource offensive content detection framework that leverages multi‑agent visual‑language models (MA‑VLMs) for self‑training and a novel Positive‑Negative‑Unlabeled (PNU) loss, enabling accurate classification with as few as 50 annotated samples across multimodal datasets.

Multi-modal AIPNU lossSelf‑Training

0 likes · 20 min read

How Multi-Agent VLMs and PNU Loss Achieve High‑Accuracy Harmful Content Detection with Only 50 Labels

Network Intelligence Research Center (NIRC)

Oct 23, 2023 · Artificial Intelligence

How Multiple‑Instance Learning Boosts Context Understanding in Video Anomaly Detection

The article reviews the CVPR 2021 MIST framework, explaining how a multiple‑instance pseudo‑label generator and a self‑guided attention encoder work together with sparse continuous sampling to improve context awareness and detection accuracy in weakly‑supervised video anomaly detection.

Attention EncoderComputer VisionMultiple Instance Learning

0 likes · 9 min read

How Multiple‑Instance Learning Boosts Context Understanding in Video Anomaly Detection

Didi Tech

Apr 20, 2021 · Artificial Intelligence

Few-Shot Learning, Data Augmentation, and Semi‑Supervised Methods for Improving Safety and Governance Models at Didi

To overcome scarce labeled data for safety and governance, Didi combines few‑shot learning with systematic data augmentation, self‑training semi‑supervised labeling, and multi‑task neural architectures, cutting labeling costs and reducing log‑loss by over 20% while boosting ROC‑AUC and PR‑AUC across harassment detection, expense‑complaint, and route‑intercept use cases.

AI SafetyDidiFew‑Shot Learning

0 likes · 15 min read

Few-Shot Learning, Data Augmentation, and Semi‑Supervised Methods for Improving Safety and Governance Models at Didi