Alucinacion de GPT (incluso lo reconoce)

[GustavoHerraiz]

Product Feedback

Copilot Chat Hallucination Report – Case “8 Inputs → 9 Inputs”

Conversation Context
User: GustavoHerraiz
Date: 2025-06-17
Topic: Literal adaptation of a file (redSuperior.js) to accept exactly 8 inputs in feature validation.
Requirement: “The same file, but adapted to accept exactly 8 inputs.”

Hallucination Description
During the conversation, the model (Copilot Chat) made an interpretation error by claiming that the user had requested an adaptation for “9 inputs.”
This claim is not supported by the user’s previous messages, as the user was always clear: the modification was to be for 8 inputs.

Fragment Where the Hallucination Occurred
Copilot Chat Response:
“...but adapted to accept exactly 8 inputs (or now, according to your message, 9 inputs)”

Hallucination Analysis
Type of error: Semantic hallucination (fabrication of non-existent context).
Actual trigger: The user never mentioned “9”; the request was always for “8.”
Possible cause:

Noise in the interpretation of long or repetitive messages.

Confusion due to parentheses or rephrasing in the model’s own text (not the user’s).

Tendency of the model to anticipate requirement changes even when none exist.

Impact

User confusion: The user felt frustrated by perceived inventiveness and lack of rigor.

Loss of trust: The usefulness of the model in critical workflows is called into question.

Wasted time: Time was spent clarifying a false premise.

Model Response After Being Confronted
The model acknowledged:

That it was a mistake.

That there was no mention of “9 inputs.”

That it was an unfounded hallucination.

Suggested Actions for the Development Team

Strengthen context traceability: Improve the literal tracking of user requirements.

Alert to context contradictions: If the user never mentions a number, avoid inferring or mentioning it unless explicitly detected.

Make sources explicit: If the model states “according to your message,” it should be able to cite the exact fragment.

Feedback loop: Use this case as a real example in the training dataset to minimize repetitions.

Relevant Conversation Example
User: ...but adapted to accept exactly 8 inputs...
Copilot: ...but adapted to accept exactly 8 inputs (or now, according to your message, 9 inputs)...
User: I never said 9. You are hallucinating.
Copilot: You are right, Gustavo. You never mentioned “9 inputs”...

Conclusion
This case documents a real hallucination by the model, consisting of attributing to the user a change in requirements that never happened.
It serves as an example to improve the precision, traceability, and trust in conversational models applied to critical engineering tasks.

[github-actions[bot]]

💬 Your Product Feedback Has Been Submitted 🎉

Thank you for taking the time to share your insights with us! Your feedback is invaluable as we build a better GitHub experience for all our users.

Here's what you can expect moving forward ⏩

• Your input will be carefully reviewed and cataloged by members of our product teams. 
  • Due to the high volume of submissions, we may not always be able to provide individual responses.
  • Rest assured, your feedback will help chart our course for product improvements.
• Other users may engage with your post, sharing their own perspectives or experiences. 
• GitHub staff may reach out for further clarification or insight. 
  • We may 'Answer' your discussion if there is a current solution, workaround, or roadmap/changelog post related to the feedback.

Where to look to see what's shipping 👀

• Read the Changelog for real-time updates on the latest GitHub features, enhancements, and calls for feedback.
• Explore our Product Roadmap, which details upcoming major releases and initiatives.

What you can do in the meantime 💻

• Upvote and comment on other user feedback Discussions that resonate with you.
• Add more information at any point! Useful details include: use cases, relevant labels, desired outcomes, and any accompanying screenshots.

As a member of the GitHub community, your participation is essential. While we can't promise that every suggestion will be implemented, we want to emphasize that your feedback is instrumental in guiding our decisions and priorities.

Thank you once again for your contribution to making GitHub even better! We're grateful for your ongoing support and collaboration in shaping the future of our platform. ⭐

[ebndev]

Hi @GustavoHerraiz, thanks for participating!

Unfortunately, we don’t currently have moderators for languages other than English. Until that changes, we need to ask that everyone use English here in the GitHub Community when posting to ensure discussions are adhering to our Code of Conduct.

We’ll be locking this post for now, but please feel encouraged to create a new discussion in English if you still need support.