mlsanityData to 7 Oct 2026

TERMon: Detecting Persistent Behavioral Threats in Edge AI via Hardware-Native Ternary Runtime Monitor

arXiv 2609.21713 ↗First posted 18 Sep 2026cs.CR cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Edge AI accelerators are increasingly deployed in safety-critical environments, where model outputs may control physical actuators, make access-control decisions, or trigger alarms. In these settings, runtime failures often remain undetected because model corruption, distribution shift, and adversarial inputs can still produce well-formed, confident predictions. This paper presents TERMon, a lightweight hardware runtime monitor that detects such anomalies by observing inference behavior rather than re-executing or formally verifying the model. TERMon represents class-conditional trusted behavior as hardware-efficient ternary patterns that are matched in parallel against a thermometer-encoded fingerprint. The ternary encoding reproduces the corresponding unquantized range decision exactly. TERMon detects harmful weight corruptions in proportion to their behavioral impact, while out-of-distribution and adversarial inputs are largely not separable using the monitored features at a strict false-positive operating point. We implemented TERMon on a PYNQ-Z2 FPGA, and the pipelined design requires no on-chip block RAM or DSPs and has a two-cycle decision latency.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 16 of its 31 references that are in the map. First posted on arXiv 18 Sep 2026.

Builds on

16 of 31 references.

References · 16

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2017On Calibration of Modern Neural Networkscited by 4,860
  3. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  4. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,362
  5. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,472
  6. 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,274
  7. 2017Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methodscited by 542
  8. 2017Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networkscited by 484
  9. 2019Bit-Flip Attack: Crushing Neural Network with Progressive Bit Searchcited by 106
  10. 2019A Survey of Deep Learning Applications to Autonomous Vehicle Controlcited by 101
  11. 2019Terminal Brain Damage: Exposing the Graceless Degradation in Deep Neural Networks Under Hardware Fault Attackscited by 55
  12. 2019Outside the Box: Abstraction-Based Monitoring of Neural Networkscited by 36
4 more
  1. 2018Analyzing and Mitigating the Impact of Permanent Faults on a Systolic Array Based Neural Network Acceleratorcited by 25
  2. 2021RADAR: Run-time Adversarial Weight Attack Detection and Accuracy Recoverycited by 18
  3. 2021HASHTAG: Hash Signatures for Online Detection of Fault-Injection Attacks on Deep Neural Networkscited by 9
  4. 2021Towards a Safety Case for Hardware Fault Tolerance in Convolutional Neural Networks Using Activation Range Supervisioncited by 7

Built on by

Not cited yet.