Namely, the TOS violations that your users have been hit with seem to fail a criteria of explanability due to the broadly subjective nature of the articles of the TOS that they appear to be violating. Both Erin Biba's short-term and Link's long term ban demonstrate how your TOS fails in this regard.
October 7, 2025 at 8:53 PM
Namely, the TOS violations that your users have been hit with seem to fail a criteria of explanability due to the broadly subjective nature of the articles of the TOS that they appear to be violating. Both Erin Biba's short-term and Link's long term ban demonstrate how your TOS fails in this regard.
Still further, because the reasons for moderation did not (in my view) explain the nature of the violations, it becomes difficult to see how your moderation decisions satisfy your criterion for explanability. That is, it seems as though your moderation decisions are based on vibes and not evidence.
October 7, 2025 at 8:53 PM
Still further, because the reasons for moderation did not (in my view) explain the nature of the violations, it becomes difficult to see how your moderation decisions satisfy your criterion for explanability. That is, it seems as though your moderation decisions are based on vibes and not evidence.
Obviously, this has direct implications for "prompt engineering" (and we clearly need better explanability tools for this). Slides from a precedent interpretability research from Petar Veličković.
August 24, 2025 at 4:35 PM
Obviously, this has direct implications for "prompt engineering" (and we clearly need better explanability tools for this). Slides from a precedent interpretability research from Petar Veličković.
I'll be giving an invited talk at The First Workshop on the Application of LLM Explainability to Reasoning and Planning at COLM'2025
xllm-reasoning-planning-workshop.github.io
Explanability? Reasoning? Planning? Language Models? Checks all the boxes for me!
xllm-reasoning-planning-workshop.github.io
Explanability? Reasoning? Planning? Language Models? Checks all the boxes for me!
XLLM-Reason-Plan
Website for the Workshop on the Application of LLM Explainability to Reasoning and Planning at COLM 2025
xllm-reasoning-planning-workshop.github.io
June 3, 2025 at 6:14 PM
I'll be giving an invited talk at The First Workshop on the Application of LLM Explainability to Reasoning and Planning at COLM'2025
xllm-reasoning-planning-workshop.github.io
Explanability? Reasoning? Planning? Language Models? Checks all the boxes for me!
xllm-reasoning-planning-workshop.github.io
Explanability? Reasoning? Planning? Language Models? Checks all the boxes for me!
Published at #irrj: "Exploring Embedding Interpretability by Correspondences Between Topic Models and Text Embeddings" by Meng Yuan, Lida Rashidi, and Justin Zobel. #informationretrieval, #embeddinginterpretability, #explanability, #topicmodelling
https://doi.org/10.54195/irrj.23703
https://doi.org/10.54195/irrj.23703
Exploring Embedding Interpretability by Correspondences Between Topic Models and Text Embeddings
| Information Retrieval Research
irrj.org
December 10, 2025 at 9:07 AM
Published at #irrj: "Exploring Embedding Interpretability by Correspondences Between Topic Models and Text Embeddings" by Meng Yuan, Lida Rashidi, and Justin Zobel. #informationretrieval, #embeddinginterpretability, #explanability, #topicmodelling
https://doi.org/10.54195/irrj.23703
https://doi.org/10.54195/irrj.23703
Fully Funded PhD position on Cybersecurity, Explanability, and Incident Response
Contact:
Prof. Dr. Giancarlo Guizzardi
University of Twente
Netherlands
bit.ly/BolsaGiancarlo
Contact:
Prof. Dr. Giancarlo Guizzardi
University of Twente
Netherlands
bit.ly/BolsaGiancarlo
bolsa.md
GitHub Gist: instantly share code, notes, and snippets.
bit.ly
September 26, 2024 at 9:34 PM
Fully Funded PhD position on Cybersecurity, Explanability, and Incident Response
Contact:
Prof. Dr. Giancarlo Guizzardi
University of Twente
Netherlands
bit.ly/BolsaGiancarlo
Contact:
Prof. Dr. Giancarlo Guizzardi
University of Twente
Netherlands
bit.ly/BolsaGiancarlo
Explanability reading 3:
"Two concepts of causation "
https://maxkasy.github.io/home/files/other/ML_Econ_Oxford/Ned_Hall_2_concepts_of_causation.pdf
Argues: More than one notion of causal explanation is needed.
"Counterfactual causation" vs "production."
4/5
"Two concepts of causation "
https://maxkasy.github.io/home/files/other/ML_Econ_Oxford/Ned_Hall_2_concepts_of_causation.pdf
Argues: More than one notion of causal explanation is needed.
"Counterfactual causation" vs "production."
4/5
maxkasy.github.io
September 21, 2024 at 2:30 PM
Explanability reading 3:
"Two concepts of causation "
https://maxkasy.github.io/home/files/other/ML_Econ_Oxford/Ned_Hall_2_concepts_of_causation.pdf
Argues: More than one notion of causal explanation is needed.
"Counterfactual causation" vs "production."
4/5
"Two concepts of causation "
https://maxkasy.github.io/home/files/other/ML_Econ_Oxford/Ned_Hall_2_concepts_of_causation.pdf
Argues: More than one notion of causal explanation is needed.
"Counterfactual causation" vs "production."
4/5
Yeah.... but AI companies by ethics tho?
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
April 4, 2025 at 8:05 AM
Yeah.... but AI companies by ethics tho?
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
Explanability reading 1:
"The Mythos of Model Interpretability"
https://arxiv.org/abs/1606.03490
Argues: The desire for explaining models can only be understood based if the decision-problems they solve is incompletely specified.
2/5
"The Mythos of Model Interpretability"
https://arxiv.org/abs/1606.03490
Argues: The desire for explaining models can only be understood based if the decision-problems they solve is incompletely specified.
2/5
The Mythos of Model Interpretability
Supervised machine learning models boast remarkable predictive capabilities. But can you trust your model? Will it work in deployment? What else can it tell you about the world? We want models to be…
arxiv.org
September 19, 2024 at 2:30 PM
Explanability reading 1:
"The Mythos of Model Interpretability"
https://arxiv.org/abs/1606.03490
Argues: The desire for explaining models can only be understood based if the decision-problems they solve is incompletely specified.
2/5
"The Mythos of Model Interpretability"
https://arxiv.org/abs/1606.03490
Argues: The desire for explaining models can only be understood based if the decision-problems they solve is incompletely specified.
2/5
Explanability reading 2:
"Counterfactual Explanations without Opening the Black Box"
https://arxiv.org/abs/1711.00399
Argues: To explain (algorithmic) decisions, for legal contestation or behavior change, requires asking for small(est) change of inputs that changes decision.
3/5
"Counterfactual Explanations without Opening the Black Box"
https://arxiv.org/abs/1711.00399
Argues: To explain (algorithmic) decisions, for legal contestation or behavior change, requires asking for small(est) change of inputs that changes decision.
3/5
Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR
There has been much discussion of the right to explanation in the EU General Data Protection Regulation, and its existence, merits, and disadvantages. Implementing a right to explanation that opens…
arxiv.org
September 20, 2024 at 2:30 PM
Explanability reading 2:
"Counterfactual Explanations without Opening the Black Box"
https://arxiv.org/abs/1711.00399
Argues: To explain (algorithmic) decisions, for legal contestation or behavior change, requires asking for small(est) change of inputs that changes decision.
3/5
"Counterfactual Explanations without Opening the Black Box"
https://arxiv.org/abs/1711.00399
Argues: To explain (algorithmic) decisions, for legal contestation or behavior change, requires asking for small(est) change of inputs that changes decision.
3/5
same. fortunate enough that i don't see it happening on my team. but, i've asked a couple engineers where i work lately what exactly is the purpose of sections of SQL and python code are and the lack of explanability is insane
May 13, 2026 at 6:10 PM
same. fortunate enough that i don't see it happening on my team. but, i've asked a couple engineers where i work lately what exactly is the purpose of sections of SQL and python code are and the lack of explanability is insane
Today I've completed this course from Kaggle, on Machine Learning explanability. It's a great course, including partial dependency plots, permutation importance, and SHAP values. #ai-challenge
llm-challenge www.kaggle.com/learn/machin...
llm-challenge www.kaggle.com/learn/machin...
Learn Machine Learning Explainability Tutorials
Extract human-understandable insights from any model.
www.kaggle.com
November 28, 2024 at 10:20 PM
Today I've completed this course from Kaggle, on Machine Learning explanability. It's a great course, including partial dependency plots, permutation importance, and SHAP values. #ai-challenge
llm-challenge www.kaggle.com/learn/machin...
llm-challenge www.kaggle.com/learn/machin...
arXiv:2504.21841v1 Announce Type: new
Abstract: Neural network-based policies have demonstrated success in many robotic applications, but often lack human-explanability, which poses challenges in safety-critical deployments. To address this, we [1/6 of https://arxiv.org/abs/2504.21841v1]
Abstract: Neural network-based policies have demonstrated success in many robotic applications, but often lack human-explanability, which poses challenges in safety-critical deployments. To address this, we [1/6 of https://arxiv.org/abs/2504.21841v1]
May 1, 2025 at 6:03 AM
arXiv:2504.21841v1 Announce Type: new
Abstract: Neural network-based policies have demonstrated success in many robotic applications, but often lack human-explanability, which poses challenges in safety-critical deployments. To address this, we [1/6 of https://arxiv.org/abs/2504.21841v1]
Abstract: Neural network-based policies have demonstrated success in many robotic applications, but often lack human-explanability, which poses challenges in safety-critical deployments. To address this, we [1/6 of https://arxiv.org/abs/2504.21841v1]
Interesting paper published recently. The title says it all: "Saliency Maps Give a False Sense of Explanability to Image Classifiers: An Empirical Evaluation across Methods and Metrics": openreview.net/forum?id=Hft...
openreview.net
November 20, 2024 at 7:04 AM
Interesting paper published recently. The title says it all: "Saliency Maps Give a False Sense of Explanability to Image Classifiers: An Empirical Evaluation across Methods and Metrics": openreview.net/forum?id=Hft...
AI companies by ethics:
Anthropic (Claude): Science-first, decent safety efforts (not nearly enough)
XAI (Grok): Explainable AI is at least a safer approach...
OpenAI (ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta (Llama): Chief scientist dismisses AI dangers (really)
Anthropic (Claude): Science-first, decent safety efforts (not nearly enough)
XAI (Grok): Explainable AI is at least a safer approach...
OpenAI (ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta (Llama): Chief scientist dismisses AI dangers (really)
April 4, 2025 at 8:12 AM
AI companies by ethics:
Anthropic (Claude): Science-first, decent safety efforts (not nearly enough)
XAI (Grok): Explainable AI is at least a safer approach...
OpenAI (ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta (Llama): Chief scientist dismisses AI dangers (really)
Anthropic (Claude): Science-first, decent safety efforts (not nearly enough)
XAI (Grok): Explainable AI is at least a safer approach...
OpenAI (ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta (Llama): Chief scientist dismisses AI dangers (really)
Oh know right?!
AI companies by ethics:
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
AI companies by ethics:
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
April 4, 2025 at 8:04 AM
Oh know right?!
AI companies by ethics:
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
AI companies by ethics:
Anthropic(Claude): Science-first, decent safety efforts(not enough)
XAI(Grok): Explainable AI is at least a safer approach...
OpenAI(ChatGPT): Biz-first, pseudo-explanability, plugging not solving safety
Meta(Llama): Chief scientist dismisses AI dangers
🎥 John Snow Labs – Healthcare AI Company posted a video
🏢 https://unjobs.org/organizations/john-snow-labs
🏷️ https://unjobs.org/themes/transparency
🏢 https://unjobs.org/organizations/john-snow-labs
🏷️ https://unjobs.org/themes/transparency
Deploying Privacy-Preserving Generative AI from Edge to Enterprise
Santosh Kumar, Technical Leader – AI at HCLTech USA, presents "Deploying Privacy-Preserving Generative AI from Edge to Enterprise."
Timestamps:
00:05 Intro: The Challenge of Responsible GenAI Deployment
01:41 4 Critical Deployment Challenges (Latency, Leakage, Silos, Sensitivity)
03:22 The Privacy-Preserving Architecture Overview
04:58 Pillars: Federated Learning + Differential Privacy
06:28 Real-World Healthcare AI Example (HIPAA Compliance & Trust)
08:00 XAI & Compliance: Explanability for Audits (GDPR, ISO)
09:52 Future Outlook: Multimodal & Self-Regulating AI
The rapid evolution of Generative AI is reshaping industries, yet real-world deployment demands more than just powerful models—it requires trust, scalability, and domain alignment. This talk explores how to architect and operationalize trustworthy generative AI systems across enterprise and healthcare environments, from cloud to edge.
Drawing on extensive AI leadership experience, the session unpacks practical frameworks for deploying transformer-based models, federated learning architectures, and privacy-preserving GenAI applications that align with regulatory and ethical imperatives.
Attendees will gain insights into designing secure, explainable AI systems that address real-world challenges such as electronic health record summarization, predictive diagnostics, claims optimization, and enterprise document intelligence. The talk also highlights strategies for achieving model robustness, transparency, and policy compliance—key pillars for responsible AI adoption.
Whether scaling GenAI in production or exploring edge-AI opportunities, this session provides a blueprint for building AI that is not only powerful but also purposeful.
This video is an official session recording from the Applied AI Summit 2025, hosted by John Snow Labs.
Connect with us:
Our website: https://www.johnsnowlabs.com/
LinkedIn: https://www.linkedin.com/company/johnsnowlabs
Facebook: https://www.facebook.com/JohnSnowLabsInc
X: https://x.com/JohnSnowLabs
#PrivacyPreservingAI #FederatedLearning #MedicalAI #GenerativeAI #DifferentialPrivacy #SecureAI #EnterpriseAI #ResponsibleAI #HIPPA #EdgeAI #DataSecurity
www.youtube.com
November 7, 2025 at 12:08 AM
🎥 John Snow Labs – Healthcare AI Company posted a video
🏢 https://unjobs.org/organizations/john-snow-labs
🏷️ https://unjobs.org/themes/transparency
🏢 https://unjobs.org/organizations/john-snow-labs
🏷️ https://unjobs.org/themes/transparency
There's been some stuff written on explanability and LLMs (there was a talk on it at the AAP this week) which suggests that we don't have good information about why Chatbot X does Thing Y, but in this case I think there's also the problem that tech companies bullshit about how their products work.
Elon Musk’s artificial intelligence company, xAI, said that its Grok chatbot relied too heavily on input from users of his social media platform X after a code update, causing it to share a series of antisemitic comments on Tuesday.
Grok Chatbot Mirrored X Users’ ‘Extremist Views’ in Antisemitic Posts, xAI Says
Elon Musk’s artificial intelligence company said its Grok chatbot had also undergone a code update that caused it to share antisemitic messages this week.
trib.al
July 12, 2025 at 7:45 PM
There's been some stuff written on explanability and LLMs (there was a talk on it at the AAP this week) which suggests that we don't have good information about why Chatbot X does Thing Y, but in this case I think there's also the problem that tech companies bullshit about how their products work.
Are you ready for the future of AI? 🤖💻 What's the biggest challenge in implementing #ChatGPT-like models in your projects?
A) Integration with legacy code 🤔
B) Data quality and bias concerns 📊
C) Scalability and performance 🚀
D) Explanability and transparency 🤷♂️ #AIChallenge
A) Integration with legacy code 🤔
B) Data quality and bias concerns 📊
C) Scalability and performance 🚀
D) Explanability and transparency 🤷♂️ #AIChallenge
My Linkedin
My Linkedin
www.linkedin.com
August 20, 2025 at 10:35 AM
Are you ready for the future of AI? 🤖💻 What's the biggest challenge in implementing #ChatGPT-like models in your projects?
A) Integration with legacy code 🤔
B) Data quality and bias concerns 📊
C) Scalability and performance 🚀
D) Explanability and transparency 🤷♂️ #AIChallenge
A) Integration with legacy code 🤔
B) Data quality and bias concerns 📊
C) Scalability and performance 🚀
D) Explanability and transparency 🤷♂️ #AIChallenge
🚀 What's the biggest challenge you're facing with #AI integration in your app? 🤔
A) Training data quality issues
B) Model complexity and maintenance
C) Integrating with existing architecture
D) Explanability and transparency concerns
#ArtificialIntelligence
A) Training data quality issues
B) Model complexity and maintenance
C) Integrating with existing architecture
D) Explanability and transparency concerns
#ArtificialIntelligence
My Linkedin
My Linkedin
www.linkedin.com
August 1, 2025 at 7:55 AM
🚀 What's the biggest challenge you're facing with #AI integration in your app? 🤔
A) Training data quality issues
B) Model complexity and maintenance
C) Integrating with existing architecture
D) Explanability and transparency concerns
#ArtificialIntelligence
A) Training data quality issues
B) Model complexity and maintenance
C) Integrating with existing architecture
D) Explanability and transparency concerns
#ArtificialIntelligence
What's the biggest hurdle in your #AI development journey? 🤖💻
A) Data quality and preprocessing
B) Model training and optimization
C) Integration with legacy systems
D) Explanability and bias detection
#AIDevelopment
A) Data quality and preprocessing
B) Model training and optimization
C) Integration with legacy systems
D) Explanability and bias detection
#AIDevelopment
My Linkedin
My Linkedin
www.linkedin.com
July 21, 2025 at 6:35 AM
What's the biggest hurdle in your #AI development journey? 🤖💻
A) Data quality and preprocessing
B) Model training and optimization
C) Integration with legacy systems
D) Explanability and bias detection
#AIDevelopment
A) Data quality and preprocessing
B) Model training and optimization
C) Integration with legacy systems
D) Explanability and bias detection
#AIDevelopment