aakas
@aakas888
Followers
3
Following
19
Statuses
11
RT @Ki_Seki_here: Why is inference-time scaling crucial, as @OpenAI o1 shows? LLMs learn world knowledge, but naive prompting turns them in…
0
17
0
RT @Ki_Seki_here: Sharing our latest work! *Internal Consistency and Self-Feedback in Large Language Models: A Survey* links: Hugging Face…
0
16
0
In our method, HSS stands for Hard Samples Selection, which involves selecting HSS to allow the large model to summarize and generate Rules for use by subsequent models; this method is very similar to ICPL (In-Context Principle Learning from Mistakes).
@_akhaliq This work is highly similar to the one we published a month ago, in which we propose a method named SLEICL. It involves learning from examples using strong language models and then summarizing and transferring these learned skills to weaker language models.
0
0
1
@aman_madaan @Swarooprm7 This work is highly similar to the one we published a month ago, in which we propose a method named SLEICL. It involves learning from examples using strong language models and then summarizing and transferring these learned skills to weaker language models.
0
0
0
@ai_database 🚀 Excited to share our latest research, which leverages strong language models to learn and summarize skills from examples, then transfers these to weaker models, and the small model is even able to achieve performance surpassing that of GPT-4.
0
0
0
@arankomatsuzaki 🚀 Excited to share our latest research: "Grimoire is All You Need for Enhancing Large Language Models", which leverages strong language models to learn and summarize skills from examples, then transfers these to weaker models.
0
0
2
Some weak language models even surpass the performance of GPT4-1106- preview (zero-shot) with the aid of SLEICL.
🚀 Excited to share our latest research: "Grimoire is All You Need for Enhancing Large Language Models", which leverages strong language models to learn and summarize skills from examples, then transfers these to weaker models.#LanguageModels #ICL #LLMs
0
0
3
🚀 Excited to share our latest research: "Grimoire is All You Need for Enhancing Large Language Models", which leverages strong language models to learn and summarize skills from examples, then transfers these to weaker models.#LanguageModels #ICL #LLMs
0
1
2