Skip to main content

An incomplete Goliath, Google, to launch undercooked tools

 


Google announced a slew of AI product integrations at their I/O 2023 keynote event this week. It seems that the core technology behind these will be its new PaLM2 LLM. That's a problem, as The Guardian article concluded:

In its preliminary research, the company warned that systems built on PaLM 2 “continue to produce toxic language harms”, with some languages issuing “toxic” responses to queries about black people in almost a fifth of all tests, part of the reason the Bard chatbot is only available in three languages at launch. 

Hinton wouldn't have approved. PaLM 2 will steal a march on OpenAI/Microsoft as it will be the first Multimodal GPT to be launched to the public. According the a Google blog the model will have the following capabilities:

  • Multilinguality: PaLM 2 is more heavily trained on multilingual text, spanning more than 100 languages. This has significantly improved its ability to understand, generate and translate nuanced text — including idioms, poems and riddles — across a wide variety of languages, a hard problem to solve. PaLM 2 also passes advanced language proficiency exams at the “mastery” level.
  • Reasoning: PaLM 2’s wide-ranging dataset includes scientific papers and web pages that contain mathematical expressions. As a result, it demonstrates improved capabilities in logic, common sense reasoning, and mathematics.
  • Coding: PaLM 2 was pre-trained on a large quantity of publicly available source code datasets. This means that it excels at popular programming languages like Python and JavaScript, but can also generate specialized code in languages like Prolog, Fortran and Verilog.
It seems quaint that Fortran is an included programming language. The paper Google published alongside the launch of PaLM 2 is rather opaque. It doesn't indicate how the model was trained for instance. What the paper states is PaLM 2 is trained on a dataset that includes a higher percentage of non-English data than previous large language models, which is beneficial for multilingual tasks (e.g., translation and multilingual question answering), as the model is exposed to a wider variety of languages and cultures.

  • In addition to non-English monolingual data, PaLM 2 is also trained on parallel data covering hundreds of languages in the form of source and target text pairs where one side is in English.
  • The inclusion of parallel multilingual data further improves the model’s ability to understand and generate multilingual text.
  • Even though PaLM 2 has a smaller proportion of English data than PaLM, we still observe significant improvements on English evaluation datasets, as described in Section 4.
  • PaLM 2 was trained to increase the context length of the model significantly beyond that of PaLM.
I don't know what the user base of all the Google products is, from maps to docs and search, but it's likely that more people will be exposed to Google AI tools than any other GPT, once the roll out is complete. Doing so with such an incomplete model seems a high risk strategy. 
Labels:

Comments

Post a Comment

Popular posts from this blog

The AI Dilemma and "Gollem-Class" AIs

From the Center for Humane Technology Tristan Harris and Aza Raskin discuss how existing A.I. capabilities already pose catastrophic risks to a functional society, how A.I. companies are caught in a race to deploy as quickly as possible without adequate safety measures, and what it would mean to upgrade our institutions to a post-A.I. world. This presentation is from a private gathering in San Francisco on March 9th with leading technologists and decision-makers with the ability to influence the future of large-language model A.I.s. This presentation was given before the launch of GPT-4. One of the more astute critics of the tech industry, Tristan Harris, who has recently given stark evidence to Congress. It is worth watching both of these videos, as the Congress address gives a context of PR industry and it's regular abuses. "If we understand the mechanisms and motives of the group mind, it is now possible to control and regiment the masses according to our will without their
Post a Comment
Read more

Beware the Orca, the challenge to ChatGPT and Palm2 is here

  So Google's 'we have no moat' paper was correct. If you train an LLM wisely then it's cost effective and cheap to produce a small LLM that is able to compete or even beat established, costly LLMs, as Microsoft has just found. It's another excellent video from AI Explained, who goes through some of the training procedures, which I won't get into here. Orca, is a model that learns from large foundation models (LFMs) like GPT-4 and ChatGPT by imitating their reasoning process. Orca uses rich signals such as explanations and complex instructions to improve its performance on various tasks. Orca outperforms other instruction-tuned models and achieves similar results to ChatGPT on zero-shot reasoning benchmarks and professional and academic exams. The paper suggests that learning from explanations is a promising way to enhance model skills. Smaller models are often overestimated in their abilities compared to LFMs, and need more rigorous evaluation methods. Explana
Post a Comment
Read more

What is happening inside of the black box?

  Neel Nanda is involved in Mechanistic Interpretability research at DeepMind, formerly of AnthropicAI, what's fascinating about the research conducted by Nanda is he gets to peer into the Black Box to figure out how different types of AI models work. Anyone concerned with AI should understand how important this is. In this video Nanda discusses some of his findings, including 'induction heads', which turn out to have some vital properties.  Induction heads are a type of attention head that allows a language model to learn long-range dependencies in text. They do this by using a simple algorithm to complete token sequences like [A][B] ... [A] -> [B]. For example, if a model is given the sequence "The cat sat on the mat," it can use induction heads to predict that the word "mat" will be followed by the word "the". Induction heads were first discovered in 2022 by a team of researchers at OpenAI. They found that induction heads were present in
Post a Comment
Read more