Rubrix is a production-ready Python framework for exploring, annotating, and managing data in NLP projects.

Getting started with 3D content for synthetic data (Unity)

More reviews

SBIRs

• 9:15 Standup. Not sure what to talk about here given the new schedule crazyness
  • It also occurs to me that since I’ll be adapting my academic research code to produce the demo, there’s no IP for anyone being developed for this effort.
• More poking at Svelte with Zach? Some progress. Still can’t get to switch pages
• 11:00 Kickoff meeting – looks like we have a bit more time
• 2:00 Adversarial reinforcement tagup

GPT Agents

• Need to generate new tweets from the chinavirus, covid, and sars-cov-2 models using the prompt ‘[[[‘ as a baseline to compare with the ground truth – done!
• Need to sample ground truth and put it in the gpt_experiments tables

Need to tell the shop that it’s a 2016 Promaster

More reviews

SBIRs

• Made some progress on Svelte, but still stuck on routing. Talking to Zach
• Meeting about slides. Or schedule has shrunk from 3 months to six weeks. Massive shift in plans and proposal

GPT Agents

• Go over untrained model results
• See if we can make the chess models talk about having tea with the Queen. I win!
• Need to generate new tweets from the chinavirus, covid, and sars-cov-2 models using the prompt ‘[[[‘ as a baseline to compare with the ground truth

WikiGraphs: A Wikipedia Text – Knowledge Graph Paired Dataset

• We present a new dataset of Wikipedia articles each paired with a knowledge graph, to facilitate the research in conditional text generation, graph generation and graph representation learning. Existing graph-text paired datasets typically contain small graphs and short text (1 or few sentences), thus limiting the capabilities of the models that can be learned on the data. Our new dataset WikiGraphs is collected by pairing each Wikipedia article from the established WikiText-103 benchmark (Merity et al., 2016) with a subgraph from the Freebase knowledge graph (Bollacker et al., 2008). This makes it easy to benchmark against other state-of-the-art text generative models that are capable of generating long paragraphs of coherent text. Both the graphs and the text data are of significantly larger scale compared to prior graph-text paired datasets. We present baseline graph neural network and transformer model results on our dataset for 3 tasks: graph -> text generation, graph -> text retrieval and text -> graph retrieval. We show that better conditioning on the graph provides gains in generation and retrieval quality but there is still large room for improvement.

Truck stuff – need to verify that they know it’s a 2016

Reviewing papers

SBIRs

• Continuing to work on Svelte. Trying to get previous useful lessons to show up as pages, but they are svelte files, not HTML, so I’m not sure how to point to them
• Pre-meeting
  • Scheduling. Orest wants to finish Oct 29, but we’re already a week into September, so I’m going to counter with Nov 5
  • Get slides done for Thurs meeting. Tried to get MARCOM to help with formatting, but the fuse is too short
  • Orest set up a meeting that conflicts with the GPT meeting. Trying to get him to move it, otherwise send a note that I will be about 15 min late

GPT Agents

• Go over untrained model results
• See if we can make the chess models talk about having tea with the Queen

It’s September, and after weeks of humidity and 90+ highs, a storm passed through and left ups with clear blue skies, cool nights, and beautiful days.

New article on Distill.pub! A Gentle Introduction to Graph Neural Networks

• Neural networks have been adapted to leverage the structure and properties of graphs. We explore the components needed for building a graph neural network – and motivate the design choices behind them.

Book

• Working on tweaks for today’s meeting
• 2:00 Meeting

SBIRs

• Continue with Svelte
• I seem to have been able to get typescript set up and running:

• Which gives us this:

• Work on finding a venue for the automating imagination paper
• OED Definition of imagination:
  • The power or capacity to form internal images or ideas of objects and situations not actually present to the senses, including remembered objects and situations, and those constructed by mentally combining or projecting images of previously experienced qualities, objects, and situations. Also (esp. in modern philosophy): the power or capacity by which the mind integrates sensory data in the process of perception.
• Also, using GNNs as ways of storing the relationships between the text generated by the GPT

I Asked GPT-3 About Covid-19. Its Responses Shocked Me. Generative AI systems could guide future pandemic decision-makers

• No public health authority should rely on an AI system to make recommendations, of course. But as they grow in power and reach, AI systems could become another tool in leaders’ belts, allowing them to quickly parse existing scientific knowledge for insights that could help to guide in-the-moment decision-making. As the systems become better at citing their sources and explaining their output, their value as tools for guiding decision-making will only grow, because the validity of their predictions can be checked and vetted.

SBIRs

• 7:30 Meeting with Zach. I’m going to see if he agrees with the “front-end-first” approach I’d like to try. He agrees, so I’m working my way through the tutotial
• To install a template project as per here, you have to use the git command line app

• That creates the following structure:

• Then to run the app, I use the terminal and use <ctrl> enter:

• This handles hot deployment in the browser, so I think I’m doing it right?
• This is pretty cool. Branching logic for HTML:

• And looping!

• 2:00 Meeting with Rukan & Aaron?

SBIRs

• Working with Zach to set up websocket-based project. Slow going today as we tried to figure out exactly how we want to set up the project
• Working on the getting started guide from websockets
• Developing with asyncio
• Looking more deeply at Svelte and thinking about building a standalone frontend that doesn’t interact with websockets, but fakes the functionality so that when the Python connections are added in it works?

JuryRoom

• 7:00 Meeting

So we’re officially done in Afghanistan now? One of these years, I’m going to try to figure out what the response to 9/11 cost, what the expectations were, and what actually happened

SBIRs

• Working with Zach on the webapp. We may be able to do all this with websockets and no server
• Sprint planning – done
• Starting on websockets. Installed websockets. I installed asyncio, but it’s part of Python. That’s nice! Uninstalled and everything still works
• The hello world works!

• Took a detour down SSL and got stuck on cert format issues? Look at that later
• Sending data to the browser:

That works too!

GPT-Agents

• Still cranking on generating reviews with the untrained model
• 3:00 Meeting. Made a bet with Shimei that the 800k chess model has forgotten that the Queen could drink tea. We’ll see if we can prompt the model to talk about something other than chess next week

If you want to summarize your research in a sentence… have an AI do it. SciTLDR sums up papers given an abstract, intro & conclusion. And it works impressively well: https://scitldr.apps.allenai.org (Via Twitter)

The Devil is in the Detail: Simple Tricks Improve Systematic Generalization of Transformers

• Recently, many datasets have been proposed to test the systematic generalization ability of neural networks. The companion baseline Transformers, typically trained with default hyper-parameters from standard tasks, are shown to fail dramatically. Here we demonstrate that by revisiting model configurations as basic as scaling of embeddings, early stopping, relative positional embedding, and Universal Transformer variants, we can drastically improve the performance of Transformers on systematic generalization. We report improvements on five popular datasets: SCAN, CFQ, PCFG, COGS, and Mathematics dataset. Our models improve accuracy from 50% to 85% on the PCFG productivity split, and from 35% to 81% on COGS. On SCAN, relative positional embedding largely mitigates the EOS decision problem (Newman et al., 2020), yielding 100% accuracy on the length split with a cutoff at 26. Importantly, performance differences between these models are typically invisible on the IID data split. This calls for proper generalization validation sets for developing neural networks that generalize systematically. We publicly release the code to reproduce our results.

SBIRs

• Got the client communicating with the server using Websockets and the server relaying those messages to RabbitMQ!

• Sprint Demos and story writing today
• Starting to look at Docker for this effort

GPT Agents

• Finish 1-5 star parser and start run on GPT-large, then GPT. Curious what we’ll get
  • Verified that everything seems to be working on a small run. Lots of parsing to get star values
  • Tring a full-sized run of 100 batches of 10 experiments with 10 return sequences
• OpenAI: The fine-tuning endpoint is now ready, and we’re excited to share it with you! Here’s how to get started: link

ETA Prediction with Graph Neural Networks in Google Maps

• Travel-time prediction constitutes a task of high importance in transportation networks, with web mapping services like Google Maps regularly serving vast quantities of travel time queries from users and enterprises alike. Further, such a task requires accounting for complex spatiotemporal interactions (modelling both the topological properties of the road network and anticipating events — such as rush hours — that may occur in the future). Hence, it is an ideal target for graph representation learning at scale. Here we present a graph neural network estimator for estimated time of arrival (ETA) which we have deployed in production at Google Maps. While our main architecture consists of standard GNN building blocks, we further detail the usage of training schedule methods such as MetaGradients in order to make our model robust and production-ready. We also provide prescriptive studies: ablating on various architectural decisions and training regimes, and qualitative analyses on real-world situations where our model provides a competitive edge. Our GNN proved powerful when deployed, significantly reducing negative ETA outcomes in several regions compared to the previous production baseline (40+% in cities like Sydney).
• I think that the GNNs should be usable to produce the maps themselves. Need to try this with simulation

Created a folder for Graph Neural Network research

Ride down to DC today for this and hopefully not get wet!

Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learning

• Isaac Gym offers a high performance learning platform to train policies for wide variety of robotics tasks directly on GPU. Both physics simulation and the neural network policy training reside on GPU and communicate by directly passing data from physics buffers to PyTorch tensors without ever going through any CPU bottlenecks. This leads to blazing fast training times for complex robotics tasks on a single GPU with 2-3 orders of magnitude improvements compared to conventional RL training that uses a CPU based simulator and GPU for neural networks. We host the results and videos at this https URL and isaac gym can be downloaded at this https URL.

SBIRs

• 1:00 meeting with Rukan
• Write some on the paper
• Do slides for demos
  • Add ‘assist Steve’ story
• Update repo and switch to dev. Verify that everything still works – it does! And receives messages as well. Oddly it seems to b e splitting the messages between the Python and TypeScript listeners:

GPT Agents

• Make some spreadsheets that compare the stars/sentiment properties of the relative models. Done. The models are remarkably stable, even down to 3k. They make more mistakes with the specific meta training but that seems to be about it?
• Trying to generate reviews from the untrained gpt2 models. The 117M model was (probably?) too small, so I’m trying the 774M model without finetuning. It requires two passes – the first creates the review (using a bigger prompt), and then I use the result and tack on “{}. I give it a star rating of“. Then I need to parse the ratings, which can be numbers or strings. I’ve kind of run out of energy so I’ll finish later.
• Start trying to figure out a posterior test?

SBIRs

• Continuing to make good progress with Zach on the webapp
• Here are a LOT of resources on self-play and reinforcement learning
  • How to build your own AlphaZero AI using Python and Keras
  • AlphaGo cheat sheet
  • AlphaGo Zero – How and Why it Works
  • How To Build Your Own AI To Play Any Board Game
  • MuZero: The Walkthrough (Part 1, Part 2, Part 3)

GPT Agents

• Brought all the model outputs from LIWC (manual here) and put them into a single spreadsheet. All the models are surprisingly stable, except for word count (WC):

• Here are some values beyond WC:
  • Analytical thinking — a high number reflects formal, logical, and hierarchical thinking; lower numbers reflect more informal, personal, here-and-now, and narrative thinking
  • Clout — a high number suggests that the author is speaking from the perspective of high expertise and is confident; low Clout numbers suggest a more tentative, humble, even anxious style.
  • Authentic — higher numbers are associated with a more honest, personal, and disclosing text; lower numbers suggest a more guarded, distanced form of discourse.
  • Emotional tone — a high number is associated with a more positive, upbeat style; a low number reveals greater anxiety, sadness, or hostility. A number around 50 suggests either a lack of emotionality or different levels of ambivalence

GPT Agents

• https://www.pnas.org/authors/submitting-your-manuscript – set up a paper repo in Overleaf and start to rough out
• Need to get the spreadsheets built for the 3k and 6k models. Done. This is pretty cool:

• Build a spreadsheet (and template?) for the LWIWC data

SBIRs

• 7:30 meeting with Zach. Good progress. We’re almost using RabbitMQ to talk between server-side TypeScript and server-side Python. And we made a cool diagram

JuryRoom

• 7:00 meeting

Learning to predict the cosmological structure formation

• Matter evolved under the influence of gravity from minuscule density fluctuations. Nonperturbative structure formed hierarchically over all scales and developed non-Gaussian features in the Universe, known as the cosmic web. To fully understand the structure formation of the Universe is one of the holy grails of modern astrophysics. Astrophysicists survey large volumes of the Universe and use a large ensemble of computer simulations to compare with the observed data to extract the full information of our own Universe. However, to evolve billions of particles over billions of years, even with the simplest physics, is a daunting task. We build a deep neural network, the Deep Density Displacement Model (D3M3), which learns from a set of prerun numerical simulations, to predict the nonlinear large-scale structure of the Universe with the Zel’dovich Approximation (ZA), an analytical approximation based on perturbation theory, as the input. Our extensive analysis demonstrates that D3MD3M outperforms the second-order perturbation theory (2LPT), the commonly used fast-approximate simulation method, in predicting cosmic structure in the nonlinear regime. We also show that D3MD3M is able to accurately extrapolate far beyond its training data and predict structure formation for significantly different cosmological parameters. Our study proves that deep learning is a practical and accurate alternative to approximate 3D simulations of the gravitational structure formation of the Universe.

GPT-Agents

• Generating content for the small-corpora models. 6k is done, working on 3k done
• Generated sentiment
• Do this to speed up the load of a mysql database (via stackoverflow)

mysql> use db_name;

mysql> SET autocommit=0 ; source the_sql_file.sql ; COMMIT ;

• 3:00 Meeting
• https://www.pnas.org/authors/submitting-your-manuscript – set up a paper repo in Overleaf and start to rough out
• Need to get the spreadsheets built for the 3k and 6k models
• Build a spreadsheet (and template?) for the LWIWC data
• Sent Shimei reviews from the 50k, 25k, 12k, 6k, and 3k models
• One of the really observable results is that the model tends to amplify the number of items that exist in larger quantities in the training corpora and reduce the number of items that are less common in the corpora. However, the tokens within a review seem to be unchanged. The average number of stars associated with a POSITIVE or NEGATIVE review seem very resilient.

SBIRs

• Meeting with Zach – nope, need to delay until tomorrow
• Trying to get RabbitMQ working
• Had to install Erlang from here: www.erlang.org
• Installed RabbitMQ from here: www.rabbitmq.com/install-windows.html
• Doing the Python tutorial from here: rabbitmq.com/tutorials/tutorial-one-python.html
• Seems to be working?

• Writing the consumer
• That’s working too!

• Seems plenty speedy when batched up, too

• 9:15 standup
• 1:00 Meeting about the sim for ARL. Going to talk about missile command, where the physics are simple, but the tactics are difficult.

Book

• Clean up chapter thumbnails. Done!

SBIR(s)

• Was getting started with Zach and then lost the power from about 8:30 to 1:30
• Looking into RabbitMQ
• Finishing up the NASA initial writeup

GPT Agents

• Based on the good results, trying a 6k and 3k models just to see how small we can get
• Trained up in less than 30 minutes! Generating content now

Need to look at this article in Science that does some multidimensional similarity mapping between COVID-19 variants.

• Derek Smith, an evolutionary biologist at the University of Cambridge, has worked for decades on visualizing immune evasion in the influenza virus in so-called antigenic maps. The farther apart two variants are on Smith’s maps, the less well antibodies against one virus protect against the other. In a recently published preprint, Smith’s group, together with David Montefiori’s group at Duke University, has applied the approach to mapping the most important variants of SARS-CoV-2

The Geometry of Shape Space: Application to Influenza

• Shape space was proposed over 20 years ago as a conceptual formalism in which to represent antibody/antigen binding. It has since played a key role in computational immunology. Antigens and antibodies are considered to be points in an abstract “shape space”, where coordinates of points in this space represent generalized physico-chemical properties associated with various (unspecified) physical properties related to binding, such as geometric shape, hydrophobicity, charge, etc. Distances in shape space between points representing antibodies and (the shape complement) of antigens are assumed to be related to their affinity, with small distances corresponding to high affinity.
• In this paper, we provide algorithms, related to metric and ordinal multidimensional scaling algorithms first developed in the mathematical psychology literature, which construct explicit, quantitative coordinates for points in shape space given experimental data such as hemagglutination inhibition assays, or other general affinity assays. Previously, such coordinates had been conceptual constructs and totally implicit. The dimension of shape space deduced from hemagglutination inhibition assays for influenza is low, approximately five dimensional.
• The deduction of the explicit geometry of shape space given experimental affinity data provides new ways to quantify the similarity of antibodies to antibodies, antigens to antigens, and the affinity of antigens to antibodies. This has potential utility in, e.g. strain selection decisions for annual influenza vaccines, among other applications. The analysis techniques presented here are not restricted to the analysis of antibody–antigen interactions and are generally applicable to affinity data resulting from binding assays.

SBIR(s)

• Meeting with Zach on the Webapp Framework. Made a lot of progress, though I only kind of know what’s going on. We were able to access MySQL on the server and add a D3 chart:

• Working on the NASA proposal

GPT Agents

• Make spreadsheets for other models and compare to 100k