[KongchangAI]
PaperLogCDM检测器

LogCDM

A lightweight log anomaly detection method that combines three signals: causal Transformer surprisal, graph attention structural distance, and transition-drift chi-square divergence. It fuses them through sigmoid gating and supports per-edge interpretable alerts and real-time deployment on CPU.

Timeline (last 90 days)

Nov 10

LogCDM uses a small detector with only 8.1-8.8M parameters and completes scoring of a single session on CPU in 5.5-10.5 milliseconds.

Unverified50%
Nov 10

LogCDM achieves an F1 score of 0.984 on the HDFS dataset.

Unverified50%
Nov 10

LogCDM achieves an F1 score of 0.870 on the chronologically split BGL dataset.

Unverified50%
Nov 10

LogCDM achieves an F1 score of 0.850 on the Thunderbird dataset.

Unverified50%
Nov 10

Under a unified evaluation protocol, an 8.1B-parameter LLM detector's F1 score differs from LogCDM's by only 0.013-0.12.

Unverified50%
Nov 10

On the BGL dataset, LogCDM outperforms all compact baselines.

Unverified50%
Nov 10

The core design of LogCDM is to feed three complementary signals (next-event surprisal, graph attention embeddings with reference graph distance, and transition drift statistics) to a sigmoid gate for fusion.

Unverified50%
Nov 10

The paper proves that LogCDM's transition drift statistic is exactly the frequency-weighted chi-square divergence between transition kernels.

Unverified50%
Nov 10

Under a shared-covariance Gaussian model, LogCDM's sigmoid gate is equivalent to a Bayesian posterior.

Unverified50%
Nov 10

Ablation experiments show that the structural signal brings F1 gains of 0.025-0.030 on BGL and Thunderbird, but contributes almost nothing on HDFS.

Unverified50%

1 more timeline events

All Facts (11)

Source Articles