Conversation

Notices

Embed this notice
Christine Lemmer-Webber (cwebber@social.coop)'s status on Thursday, 23-Jul-2026 04:00:35 JST Christine Lemmer-Webber
in reply to
- allison
@aparrish Well that last sentence is the goal of https://people.ucsc.edu/~lgilpin/publication/dissertation/ and I think she was on the right track
In conversation about 3 months ago from social.coop permalink
Attachments
1. No result found on File_thumbnail lookup.
  
  Anomaly Detection Through Explanations | Leilani H. Gilpin
  
  from Leilani H. Gilpin
  
  Under most conditions, complex machines are imperfect. When errors occur, as they inevitably will, these machines need to be able to (1) localize the error and (2) take appropriate action to mitigate the repercussions of a possible failure. My thesis con- tributes a system architecture that reconciles local errors and inconsistencies amongst parts. I represent a complex machine as a hierarchical model of introspective sub- systems working together towards a common goal. The subsystems communicate in a common symbolic language. In the process of this investigation, I constructed a set of reasonableness monitors to diagnose and explain local errors, and a system- wide architecture, Anomaly Detection through Explanations (ADE), which reconciles system-wide failures. The ADE architecture contributes an explanation synthesizer that produces an argument tree, which in turn can be backtracked and queried for support and counterfactual explanations. I have applied my results to explain incor- rect labels in semi-autonomous vehicle data. A series of test simulations show the accuracy and performance of this architecture based on real-world, anomalous driving scenarios. My work has opened up the new area of explanatory anomaly detection, towards a vision in which: complex machines will be articulate by design; dynamic, internal explanations will be part of the design criteria, and system-level explanations will be able to be challenged in an adversarial proceeding.
- Embed this notice
  allison (aparrish@friend.camp)'s status on Thursday, 23-Jul-2026 04:00:36 JST allison
  
  @cwebber i think the issue is that i tend to think of those criteria as categorical, not as shades of gray. like i only see value in an ML model if its training data is documented, fully understood, and gathered with consent, full stop—just inching away from "plagiarism machine" doesn't help. likewise, i have no use for a "more accurate" statistical model—I only see value in a model if i can fully understand and contextualize its predictions
  
  In conversation about 3 months ago permalink

Public

Conversation

Notices

Feeds