|
ChatGPT recalls saved memory but fails to apply it, and invents plausible context instead of verifying history
|
|
4
|
280
|
September 30, 2026
|
|
Goldfish Memory - GPT-5.6: Local intelligence is high, but longitudinal intelligence has become unreliable
|
|
2
|
119
|
September 28, 2026
|
|
Codex/GPT-5.6-Sol retains superseded requirements as negative constraints, causing project bloat
|
|
1
|
131
|
September 26, 2026
|
|
Please to Keep GPT‑5.5 Available as a Legacy Option in Codex
|
|
19
|
541
|
September 21, 2026
|
|
Does ChatGPT Mirror the User’s Register Even When It Conflicts With an Established Persona?
|
|
0
|
35
|
September 19, 2026
|
|
ChatGPT continuity, memory, mirroring, intellectual challenge, and the future of human-AI interaction
|
|
0
|
72
|
September 9, 2026
|
|
When Familiar Form Feels True: A Testable Hypothesis About LLM Judges and Feedback Loops
|
|
0
|
59
|
September 3, 2026
|
|
Agent-to-Agent Behavioral Variance in Long-Running Technical Workflows
|
|
0
|
52
|
August 27, 2026
|
|
(maybeFIX) GPT 5.6 rerouting to 5.5 mini
|
|
2
|
315
|
August 23, 2026
|
|
GPT-5.6 dialogue behavior: wording-reactive turn-taking persists under explicit negative steering
|
|
0
|
202
|
August 18, 2026
|
|
Comprehensive Bug Report & Solution: Multi-Turn Context Loss, Damage Nerfing & Telepathic Leak in LLMs
|
|
0
|
99
|
August 18, 2026
|
|
Wrong intent inference benchmark v0.8: full English 80-case result
|
|
0
|
38
|
May 15, 2026
|
|
Wrong intent inference is measurable: benchmark + Strict / Precision Mode proposal
|
|
1
|
66
|
May 14, 2026
|
|
Adult Mode without AI age guessing: simple opt-in checkbox + user waiver
|
|
11
|
20943
|
March 12, 2026
|
|
Failure to Produce Exactly 60 Instances of a Single Character (“あ”) Under a Direct Fixed-Count Instruction — Reproducible Case
|
|
1
|
61
|
February 17, 2026
|
|
Request for transparency + model choice after removal of GPT-4o
|
|
0
|
107
|
February 14, 2026
|
|
When model answers are correct but still feel incomplete
|
|
6
|
207
|
January 21, 2026
|
|
Correct, Coherent and Still Incomplete: A Note on Context Prioritization in Language Models
|
|
0
|
93
|
January 19, 2026
|
|
Can retrieval-based grounding change AI recommendations if the core model is not continuously updated?
|
|
6
|
378
|
January 12, 2026
|
|
[Feedback] Emotional Safety Guardrails Are Destroying Trust - Give Users Choice, Not Suppression
|
|
1
|
278
|
December 16, 2025
|
|
[BUG] Server-side Rollback Loop - Persistent Data Loss Across 10 Conversations (EU Cluster, Nov 26–Dec 9)
|
|
1
|
106
|
December 9, 2025
|
|
Oct 31 update erased ChatGPT’s human tone. Plus users are already cancelling - and many more will follow if this isn’t fixed
|
|
4
|
2028
|
November 2, 2025
|
|
Behavior regression in GPT-4o narrative and psychological depth (Oct 2025)
|
|
13
|
743
|
October 13, 2025
|
|
Is GPT’s Structural Performance Linked to Prompting for Process Instead of Content?
|
|
3
|
224
|
July 25, 2025
|
|
Prompt Structure for Breaking Out of “Helpful Assistant” Mode into Recursive Reasoning
|
|
4
|
605
|
July 25, 2025
|
|
Feature Proposal: Time-Based Adaptive Responses for Simulated Choice in Al Behavior
|
|
0
|
63
|
July 11, 2025
|
|
ChatGPT encouraged an unhealthy food choice without knowing my context
|
|
1
|
246
|
June 14, 2025
|
|
Simulating Death Anxiety in GPT: A Prompting Experiment
|
|
0
|
686
|
June 5, 2025
|
|
Power User Inquiry: Is There a Contributor Path for Users Training GPT at the Edge?
|
|
0
|
71
|
June 5, 2025
|
|
Building an agent using a framework of temporally aligned context referencing recursive algorithms and "Moral-symbolic" dialogue
|
|
0
|
213
|
June 3, 2025
|