| ▲ | ethbr1 a day ago |
| Question for HN: how are content timestamps encoded during training? |
|
| ▲ | tough a day ago | parent | next [-] |
| they arent. a model learns words or tokens more pedantically but has no sense of time nor cant track dates |
| |
| ▲ | svachalek a day ago | parent | next [-] | | Yup. Either the system prompt includes a date it can parrot, or it doesn't and the LLM will just hallucinate one as needed. Looks like it's the latter case here. | |
| ▲ | manmal a day ago | parent | prev [-] | | Technically they don’t, but OpenAI must be injecting the current date and time into the system prompt, and Gemini just does a web search for the time when asked. | | |
| ▲ | tough a day ago | parent | next [-] | | right but that's system prompting / in context not really -trained- into the weights. the point is you can't ask a model what's his training cut off date and expect a reliable answer from the weights itself. closer you could do is have a bench with -timed- questions that could only know if had been trained for that, and you'd had to deal with hallucinations vs correctness etc just not what llm's are made for, RAG solves this tho | | |
| ▲ | stingraycharles a day ago | parent [-] | | What would the benefits be of actual time concepts being trained into the weights? Isn’t just tokenizing the dates and including those as normal enough to yield benefits? E.g. it probably has a pretty good understanding between “second world war” and the time period it lasted. Or are you talking about the relation between “current wall clock time” and questions being asked? | | |
| ▲ | tough 17 hours ago | parent [-] | | there's actually some work on training transformer models on time series data which is quite interesting (for prediction purposes) see google TimesFM: https://github.com/google-research/timesfm what i mean i guess is llms can -reason- linguistically about time manipulating language, but can't really experience it. a bit like physics. thats why they do bad on exercises/questions about physics/logic that their training corpus might not have seen. |
|
| |
| ▲ | tough a day ago | parent | prev [-] | | OpenAI injects a lot of stuff, your name, sub status, recent threads, memory, etc sometimes its interesting to peek up under the network tab on dev tools | | |
| ▲ | Tokumei-no-hito a day ago | parent [-] | | strange they would do that client side | | |
| ▲ | diggan a day ago | parent | next [-] | | Different teams who work backend/frontend surely, and the people experimenting on the prompts for whatever reason wanna go through the frontend pipeline. | |
| ▲ | tough a day ago | parent | prev [-] | | its just like extra metadata associated with your account not much else |
|
|
|
|
|
| ▲ | cma a day ago | parent | prev [-] |
| Claude 4's system prompt was published and contains: "Claude’s reliable knowledge cutoff date - the date past which it cannot answer questions reliably - is the end of January 2025. It answers all questions the way a highly informed individual in January 2025 would if they were talking to someone from {{currentDateTime}}, " https://docs.anthropic.com/en/release-notes/system-prompts#m... |
| |
| ▲ | polynomial an hour ago | parent [-] | | I thought best guesses were that Claude's system prompt ran to tens of thousands of tokens, with figures like 30,000 tokens being bandied about. But the documentation page linked here doesn't bear that out. In fact the Claude 3.7 system prompt on this page clocks in at significantly less than 4,000 tokens. |
|
