Introducing Epic’s version control system: Lore! Built from scratch and open-sourced, we made scalability and performance core tenets from day one.
We’re releasing Lore as open source today.All source code, along with documentation of data formats and protocols are released under an MIT license and free to use. Help shape it into something even more powerful, together: https://t.co/zplVMjkRof
GPT 5.3 Codex Spark got stuck in a compaction loop
After context compaction, it seems to be repeatedly entering a loop:
"I'm doing one more audit pass"
"I'm checking a few more files"
"I'll summarize findings shortly"
@thsottiaux@jxnlco
Then context compacts again and it restarts the same audit plan.
Result:
~36 minutes of runtime
dozens of file reads/searches
multiple compactions
never produced the actual findings
Frontier LLMs are doing too much when it comes to editing code.
I'm excited to share this work on the Over-Editing problem which refers to models modifying code beyond what is asked of them.
The main findings are:
- Many frontier models Over-Edit with GPT 5.4 being the biggest culprit
- Reasoning models have a higher natural tendency to Over-Edit compared to their non-reasoning counterparts
- RL is the best approach to train models to perform minimal code editing while preventing catastrophic forgetting compared to SFT, DPO and Rejection Sampling.
Blog and details below!
“Pressure is a Privilege. And if you’re feeling any pressure or weight of expectation, you are breathing rare air, that very few of us get to live inside”.
Unbelievable piece of poetry .