Especially with Gemini Pro when providing long form textual references, providing many documents in a single context windows gives worse answers than having it summarize documents first, ask a question about the summary only, then provide the full text of the sub-documents on request (rag style or just simple agent loop).
Similarly I've personally noticed that Claude Code with Opus or Sonnet gets worse the more compactions happen, it's unclear to me whether it's just the summary gets worse, or if its the context window having a higher percentage of less relevant data, but even clearing the context and asking it to re-read the relevant files (even if they were mentioned and summarized in the compaction) gives better results.
Gemini loses coherence and reasoning ability well before the chat hits the context limitations, and according to this report, it is the best model on several dimensions.
Long story short: Context engineering is still king, RAG is not dead
Yep, it can decohere really badly with bigger context. It's not only context related though. Sometimes it can lose focus early on in a way that is impossible to get it back on track.
Gemini loses the notion of context the longer its context is: I often ask it to provide a summary of our discussion for the outside world and it will reference ideas or documents without introducing them, via anaphore, as if the outside world had knowledge of the context.
Cursor lifted "Start a new chat" limitation on gemini and i'm actually now enjoying keeping longer sessions within one window, becuase it's still very reasonable at recall, but doesnt need to restate everything each time
"Compactions" are just reducing the transcript to a summary of the transcript, right? So it makes sense that it would get worse because the agent is literally losing information, but it wouldn't be due to context rot.
The thing that would signal context rot is when you approach the auto-compact threshold. Am I thinking about this right?
I feel like the optimal coding agent would do this automatically - collect and (sometimes) summarize the required parts of code, MCP responses, repo maps etc., then combine the results into a new message in a new 'chat' that would contain all the required parts and nothing else. It's basically what I already do with aider, and I feel the performance (in situations with a lot of context) is way better than any agentic / more automated workflow I've tried so far, but it is a lot of work.
Claude Code tries, and it seems to be OK at it. It's hard to tell though and it definitely feels like sometimes you absolutely have to quit out and start again.
Have you tried NotebookLM which basically does this as an app on the bg (chunking and summarising many docs) and you can -chat- with the full corpus using RAG
Most exchanges do not reveal counter-party information smaller than the broker level. So you wouldn't know just from looking at market activity the same person causing the large futures move was also taking large options positions.
The pattern was exploitable only on the specific days that Jane Street was allegedly manipulating. How would you have figured out, without counterparty information and before noisy sales start dragging down the index, that day X is a manipulation day?
How would you have identified that there's even such a thing as a manipulation day? Do you have a model that tells you the objectively correct number of days a non-manipulated index should be lower at close?
Usually, everyone does do that, which is why only hard-to-detect patterns remain profitable. Not something obvious like "buy options in the morning and sell in the evening" as in this example.
But maybe Jane Street only traded like this on some days, so you would need to know whether they had done so before you could hope to exploit them.
I finished it in 2019~, same with skipping the Laos section that didn’t exist. I contributed a bit to Seat61 from rural local stations in Myanmar while it was still open.
Human intelligence roughly follows a normal distribution where the median is the same as the mean. In that sense OP was correct that half of the population are below average.
