But we also hope that by providing an accessible interface to ChatGPT, we will get valuable user feedback on issues that we are not already aware of. We know that many limitations remain as discussed above and we plan to make regular model updates to improve in such areas. OpenAI attempts to use its moderation API to reject requests that do not comply with its own content policies. The model would sometimes respond to inappropriate requests instead of rejecting them. Instead of responding to unclear statements with queries, ChatGPT would try to guess the user’s intent. The reasons are over-optimization and the bias of human instructors, who preferred more detailed answers in the human feedback process. Moreover, ChatGPT is too wordy, uses phrases, and repeats itself. Depending on the input, it may not answer a question, answer it incorrectly, or answer it correctly – according to OpenAI, small rephrasing can suffice. The same issue led to protests from researchers when Meta released its Galactica science model.Īccording to OpenAI, this poses a major challenge because there is no single source of truth, an overly cautious trained model rejects questions, and in supervised training, the ideal answer depends on the knowledge of the model rather than the human demonstrator.ĬhatGPT also reacts strongly to small changes in the prompts. The model can sometimes generate plausible-sounding but incorrect and nonsensical responses.
ChatGPT still has many limitationsĬommon limitations of large language models also apply to ChatGPT. All models were trained on Microsoft’s Azure AI platform. The base model of ChatGPT is a model from the GPT-3.5 series, which finished training in early 2022.
0 kommentar(er)
