465 subscribers
478 photos
23 videos
36 files
948 links
Download Telegram
😱4👌2
The False Promise of Imitating Proprietary LLMs

Abstract:

An emerging method to cheaply improve a weaker language model is to finetune it on outputs from a stronger model, such as a proprietary system like ChatGPT (e.g., Alpaca, Self-Instruct, and others). This approach looks to cheaply imitate the proprietary model's capabilities using a weaker open-source model. In this work, we critically analyze this approach. We first finetune a series of LMs that imitate ChatGPT using varying base model sizes (1.5B--13B), data sources, and imitation data amounts (0.3M--150M tokens). We then evaluate the models using crowd raters and canonical NLP benchmarks. Initially, we were surprised by the output quality of our imitation models -- they appear far better at following instructions, and crowd workers rate their outputs as competitive with ChatGPT. However, when conducting more targeted automatic evaluations, we find that imitation models close little to none of the gap from the base LM to ChatGPT on tasks that are not heavily supported in the imitation data. We show that these performance discrepancies may slip past human raters because imitation models are adept at mimicking ChatGPT's style but not its factuality. Overall, we conclude that model imitation is a false promise: there exists a substantial capabilities gap between open and closed LMs that, with current methods, can only be bridged using an unwieldy amount of imitation data or by using more capable base LMs. In turn, we argue that the highest leverage action for improving open-source models is to tackle the difficult challenge of developing better base LMs, rather than taking the shortcut of imitating proprietary systems.

> We show that these performance discrepancies may slip past human raters because imitation models are adept at mimicking ChatGPT's style but not its factuality

https://arxiv.org/pdf/2305.15717
👍1👌1
Forwarded from Axis of Ordinary
IQ scores by ethnic group in a nationally-representative sample of 10-year old American children https://humanvarieties.org/2023/05/27/iq-scores-by-ethnic-group-in-a-nationally-representative-sample-of-10-year-old-american-children/
👍5🤔21🔥1
Forwarded from Axis of Ordinary
Improving Mathematical Reasoning with Process Supervision https://openai.com/research/improving-mathematical-reasoning-with-process-supervision

"We've trained a model to achieve a new state-of-the-art in mathematical problem solving by rewarding each correct step of reasoning (“process supervision”) instead of simply rewarding the correct final answer (“outcome supervision”). In addition to boosting performance relative to outcome supervision, process supervision also has an important alignment benefit: it directly trains the model to produce a chain-of-thought that is endorsed by humans."
👍2
Forwarded from Ordo Nexus
Thanks to the independent developer, now we have browser extension Nexus Now! for Firefox, Chrome and Edge allowing you to jump from publishers' pages to STC and download the paper of interest
Holy fuck
👍5
😱3🙏1
Forwarded from daily fortunes
😱2