Positive users praise GPT-5.6's upgraded computer use, benchmarks, and instruction handling for major gains, while negative users criticize the headache of cleaning instructions, repetitive phrasing, sterile outputs, and inefficiencies.
Based on 14 visible X reactions from 53 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@yanndubs is it possible to make GPT use more normal/varied language instead of "it's not X, it's Y" in every other sentence? this is literally the most annoying thing about GPT for me
@yanndubs @thsottiaux that cleaning part sounds like a absolute headache to deal with.
@bee_human_ @yanndubs @thsottiaux Super good idea cc: @dkundel
@yanndubs @thsottiaux Pruned a lot. Big improvements.
@yanndubs is it possible to make GPT use more normal/varied language instead of "it's not X, it's Y" in every other sentence? this is literally the most annoying thing about GPT for me
@yanndubs this is so impressive
@yanndubs oss2, it’s about time
☀️Very proud of the team for training 5.6! A few of my highlights: - Great front-end aesthetics. E.g., I asked Sol to make a new blog for 5.6 with a celestial theme. - Much better CUA - Can work for much longer and is less lazy But it's not all great: - The writing quality has improved, but it's still bad in artifacts - Too many options with models/reasoning/fast mode/Cerebras/multi-agent We'll fix the above. What else should we improve? PS: belated post as there are so many other good models to train.
Tips when moving to 5.6 -Clean up old instructions. 5.6 follows them carefully, including undesirable ones buried in old skills. We had to clean many of those -Avoid Ultra if cost-constrained. Multi-agent improves evals+latency but costs much more. Or ask @thsottiaux for resets https://x.com/yanndubs/status/2076712508791128223/photo/1
Positive users praise GPT-5.6's upgraded computer use, benchmarks, and instruction handling for major gains, while negative users criticize the headache of cleaning instructions, repetitive phrasing, sterile outputs, and inefficiencies.
Based on 14 visible X reactions from 53 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@yanndubs oss2, it’s about time
☀️Very proud of the team for training 5.6! A few of my highlights: - Great front-end aesthetics. E.g., I asked Sol to make a new blog for 5.6 with a celestial theme. - Much better CUA - Can work for much longer and is less lazy But it's not all great: - The writing quality has improved, but it's still bad in artifacts - Too many options with models/reasoning/fast mode/Cerebras/multi-agent We'll fix the above. What else should we improve? PS: belated post as there are so many other good models to train.
Tips when moving to 5.6 -Clean up old instructions. 5.6 follows them carefully, including undesirable ones buried in old skills. We had to clean many of those -Avoid Ultra if cost-constrained. Multi-agent improves evals+latency but costs much more. Or ask @thsottiaux for resets https://x.com/yanndubs/status/2076712508791128223/photo/1