Issue Station
Sign In
Tools hereToolbox

codex - 💡(How to fix) Fix Model gtp5.4 xhigh fast does not follow requests, ignore them and add stupid things, lose memory like a fish [2 comments, 3 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
openai/codex#17882•Fetched 2026-04-15 06:26:55
View on GitHub
Comments
2
Participants
3
Timeline
6
Reactions
0
Author
alienwareagentalienwareagent
Participants
alienwareagentalienwareagentetraut-openaietraut-openaigithub-actions[bot]github-actions[bot]
Timeline (top)
labeled ×3commented ×2closed ×1
RAW_BUFFERClick to expand / collapse

What version of Codex CLI is running?

codex-cli 0.120.0

What subscription do you have?

Pro

Which model were you using?

gtp5.4 xhigh fast

What platform is your computer?

Ubuntu 25.10

What terminal emulator and version are you using (if applicable)?

ptyxis

What issue are you seeing?

The model is annoying wasting my time like crazy, a junior developer will not upset you and make you annoyed as the model doing lately. He eats tokens like crazy, screw my code and try to focus on the easiest things. It has been a nightmare working with lately. Please fix it make the prompt more severe and also make it follow the rules by adding things properly to it's memory. As believe me, it's acting worst than local model. I know you are working hard but it has been wasting my time and my energy. Looking forward for your fix!

What steps can reproduce the bug?

Try working with it in complex project by writing it following your rules or ask him to refactor a code in your way. He acknowledge he messed up but he keep messing up again and again and again

What is the expected behavior?

Follow the rules and not feel I am vibe coder, my rule is my rule and not him try to do it it's way and the easy way for it and burning tokens for nothing....

Additional information

No response

extent analysis

TL;DR

The issue may be resolved by adjusting the model's configuration or prompt engineering to better suit the user's needs and coding style.

Guidance

  • Review the model's documentation to understand how to properly configure it for complex projects and refactor code according to specific rules.
  • Experiment with different prompt formats and severity levels to see if it improves the model's performance and reduces token waste.
  • Consider providing more specific and detailed rules or guidelines for the model to follow, rather than relying on general prompts.
  • Evaluate the model's performance on smaller, well-defined tasks to identify potential issues with its understanding or application of the rules.

Notes

The issue description lacks technical details, making it challenging to provide a precise solution. The user's frustration with the model's performance is clear, but the root cause of the problem is unclear.

Recommendation

Apply workaround: Adjust the model's configuration and prompt engineering to better suit the user's needs, as the issue seems to be related to the model's understanding and application of the rules rather than a version-specific bug.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#tensor shape#autograd error#model save/load#optimization#mixed precision

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

codex - 💡(How to fix) Fix Model gtp5.4 xhigh fast does not follow requests, ignore them and add stupid things, lose memory like a fish [2 comments, 3 participants]