SELF-INSTRUCT: Aligning Language Model with Self Generated Instructions
The Model That Changes Everything: Alpaca Breakthrough (ft. Apple's LLM, BritGPT, Ernie and AlexaTM)
8 years of cost reduction in 5 weeks: how Stanford's Alpaca model changes everything, including the economics of OpenAI and GPT 4. The breakthrough, using self-instruct, has big implications for Apple's secret large language model, Baidu's ErnieBot, Amazon's attempts and even governmental efforts, like the newly announced BritGPT.
I will go through how Stanford put the model together, why it costs so little, and demonstrate in action versus Chatgpt and GPT 4. And what are the implications of short-circuiting human annotation like this? With analysis of a tweet by Eliezer Yudkowsky, I delve into the workings of the model and the questions it rises.
Web Demo: https://alpaca-ai0.ngrok.io/
Alpaca: https://crfm.stanford.edu/2023/03/13/alpaca.html
Ark Forecast: https://research.ark-invest.com/hubfs/1_Download_Files_ARK-Invest/Big_Ideas/ARK%20Invest_013123_Presentation_Big%20Ideas%202023_Final.pdf
Eliezer Tweet: https://twitter.com/ESYudkowsky/status/1635577836525469697
https://twitter.com/ESYudkowsky/status/1635667349792780288
Self-Instruct: https://arxiv.org/pdf/2212.10560.pdf
InstructGPT: https://openai.com/research/instruction-following
OpenAI Terms: https://openai.com/policies/terms-of-use
MMLU Test: https://arxiv.org/pdf/2009.03300.pdf
Apple LLM: https://www.nytimes.com/2023/03/15/technology/siri-alexa-google-assistant-artificial-intelligence.html
GPT 4 API: https://openai.com/pricing
Llama Models: https://arxiv.org/pdf/2302.13971.pdf
BritGPT: https://www.theguardian.com/technology/2023/mar/15/uk-to-invest-900m-in-supercomputer-in-bid-to-build-own-britgpt
Amazon: https://www.businessinsider.com/amazons-ceo-andy-jassy-on-chat-cpt-ai-2023-2?r=US&IR=T
AlexaTM: https://arxiv.org/pdf/2208.01448.pdf
Baidu Ernie: https://www.nytimes.com/2023/03/16/world/asia/china-baidu-chatgpt-ernie.html
PaLM API: https://developers.googleblog.com/2023/03/announcing-palm-api-and-makersuite.html
https://www.patreon.com/AIExplained
GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
Couairon embedding arithmetic of multimodal queries for image retrieval cvprw 2022 paper
GPT-4 Technical Report
Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models
Training Compute-Optimal Large Language Models
Triton: An Intermediate Language and Compiler for Tiled Neural Network Computations
LORA: LOW-RANK ADAPTATION OF LARGE LANGUAGE MODELS
Product quantization for nearest neighbor search
ProofNet: Autoformalizing and Formally Proving Undergraduate-Level Mathematics
LLaMA: Open and Efficient Foundation Language Models
Introducing LLaMA: A foundational, 65-billion-parameter language model
Today, we’re releasing our LLaMA (Large Language Model Meta AI) foundational model with a gated release. LLaMA is more efficient and competitive with previously published models of a similar size on existing benchmarks.
fka/awesome-chatgpt-prompts · Datasets at Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
World’s first on-device demonstration of Stable Diffusion on an Android phone
Qualcomm AI Research deploys a popular 1B+ parameter foundation model on an edge device through full-stack AI optimization.
FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
Focal Modulation Networks
CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformers
Mesa-Optimization - AI Alignment Forum
Mesa-Optimization is the situation that occurs when a learned model (such as a neural network) is itself an optimizer. In this situation, a base optimizer creates a second optimizer, called a mesa-optimizer. The primary reference work for this concept is Hubinger et al.'s "Risks from Learned Optimization in Advanced Machine Learning Systems".
Example: Natural selection is an optimization process that optimizes for reproductive fitness. Natural selection produced humans, who are themselves optimizers. Humans are therefore mesa-optimizers of natural selection.
In the context of AI alignment, the concern is that a base optimizer (e.g., a gradient descent process) may produce a learned model that is itself an optimizer, and that has unexpected and undesirable properties. Even if the gradient descent process is in some sense "trying" to do exactly what human developers want, the resultant mesa-optimizer will not typically be trying to do the exact same thing.[1]
HISTORY
Previously work under this concept was called Inner Optimizer or Optimization Daemons.
Wei Dai brings up a similar idea in an SL4 thread.[2]
The optimization daemons article on Arbital was published probably in 2016.[1]
Jessica Taylor wrote two posts about daemons while at MIRI:
* "Are daemons a problem for ideal agents?" (2017-02-11)
* "Maximally efficient agents will probably have an anti-daemon immune system" (2017-02-23)
SEE ALSO
* Inner Alignment
* Complexity of value
* Thou Art Godshatter
EXTERNAL LINKS
Video by Robert Miles
Some posts that reference optimization daemons:
* "Cause prioritization for downside-focused value systems": "Alternatively, perhaps goal preservation becomes more difficult the more capable AI systems become, in which case the future might be controlled by unstable goal functions taking turns over the steering wheel"
* "Techniques for optimizing worst-case performance": "The difficulty of optimizing worst-case performance is one of the most likely re
Mesa-Optimization is the situation that occurs when a learned model (such as a neural network) is itself an optimizer. In this situation, a base optimizer creates a second optimizer, called a mesa-optimizer. The primary reference work for this concept is Hubinger et al.'s "Risks from Learned Optimization in Advanced Machine Learning Systems".
Question Answering - OpenAI | Weaviate - vector search engine
In short
First it performs a semantic search with k=1 to find the document (e.g. a Sentence, Paragraph, Article, etc.) which is most likely to contain the answer. This step has no certainty threshold and as long as at least one document is present, it will be fetched and selected as the one most likely containing the answer. In a second step, Weaviate creates the required prompt as an input to an external call made to the OpenAI Completions endpoint. Weaviate uses the most relevant documents to establish a prompt for which OpenAI extracts the answer
Blockchained On-Device Federated Learning
New AI classifier for indicating AI-written text
We’re launching a classifier trained to distinguish between AI-written and human-written text.
We’ve trained a classifier to distinguish between text written by a human and text written by AIs from a variety of providers. While it is impossible to reliably detect all AI-written text, we believe good classifiers
Learning to Generate Reviews and Discovering Sentiment
DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature
Secure multi-party computation - Wikipedia
Secure multi-party computation (also known as secure computation, multi-party computation (MPC) or privacy-preserving computation) is a subfield of cryptography with the goal of creating methods for parties to jointly compute a function over their inputs while keeping those inputs private. Unlike traditional cryptographic tasks, where cryptography assures security and integrity of communication or storage and the adversary is outside the system of participants (an eavesdropper on the sender and receiver), the cryptography in this model protects participants' privacy from each other.
Alexis Courbet | Towards Computational Design of Self Assembling &Genetically Encodable Nanomachines
Training Your Own Dense Passage Retrieval Model | Haystack
Learn about training a Dense Passage Retrieval model and the data needed to do so.
DPR is standardly trained using a method known as in-batch negatives.
This means that positive contexts for a given query are treated as negative contexts for the other queries in the batch.
Doing so allows for a high degree of computational efficiency, thus allowing the model to be trained on large amounts of data.
Dense Passage Retrieval for Open-Domain Question Answering
Neural networks generalize because of this one weird trick - LessWrong
Produced as part of theSERI ML Alignment Theory Scholars Program- Winter 2022 Cohort …
Precise Zero-Shot Dense Retrieval without Relevance Labels