Respond to Alden’s email done

AI Coffee Break with Letitia

GOES

• Status report! Done!
• Create a new class based on utils/ScriptReaderScratch that uses the the code from least_squares_rotations.py to create data for training
• Attend the GSAW welcome and overview at 11:50 – missed it
• Create a more generic generator based on timeseriesML2\generators that will create a numpy ndarray of n-dimensional times series data. Could also use a Dataframe and have labels.
  • Randomized start, within a range
  • Adjustable noise
  • Adjustable time step
  • Different function for each row
  • Input file driven
  • Saves to csv (with a header that describes the data?) or an excel file for humans. Use the to_excel() code from EccoToXlsx for this

GPT Agents

• Run an Ecco experiment and create spreadsheets using the chess data – done

• After that, back up the gpt_experiments and commit to svn – done
• Make sure that the following are on the laptop for the 3:00 Meeting -done
  • updated gpt_experiments
  • small_feb2021
• Uploading trained models to svn. When the last one is done, zip the whole batch and put it on DropBox
• I think I know how to contribute to a project that I am not a member. I need to clone the project to my repo and work on that version. When I’m at a state that I like, then I can do a pull request. That means there are going to be one version of the source project in External and my branch in Sandboxes

I reran my monthly COVID-19 visualizations. Here’s my sample of countries. The UK is at the top of the ‘badly handled’ cluster, which includes the USA, Italy, Sweden, France and Switzerland. Germany is a bit better, and Canada really seems to be keeping things under control. The bottom cluster ranges from Finland to Senegal to China. Effective policy doesn’t seem to be related to government, wealth, population or location:

And here’s all 50 states plus territories. I switch between Republican and Democratic governors at the end. You can see that there’s not much difference except for Georgia. Something has gone horribly wrong there:

GPT Agents

• Running Ecco trend analysis with the new model that Sim made
  • I think there is a multiple embedding problem that we’ll need to address.
  • It looks really good though…

• Still training monthly models. At October 2020 now. It takes a bit under 10 hours to train most models

Pick up RV! done

GPT Agents

• Working on turning the rank matrix into a class of EccoTrendAnalytics – done!
• Need to make a ‘json’ tag for table_output and load in the ETA dict
• New version of Ecco out. I need to mergeand fold in my changes
• Still training the April model – done! On to May
• 3:30 Meeting. We played with the GPT-3 a lot

GOES

• 11:00 Meeting with Erik & Vadim. Continue working on creating data using the LS model. Architect and train a classifier. Demo the yaw flip to show capability and then focus on Nadir Point.

SBIR

• 11:30 Meeting to finalize report. Done

Book:

• 2:00 Meeting with Michelle. More pitch organization

GPT Agents

• Continuing to dig into GPT-3 prompt metaprogramming
• Now training the March model. That’s been running for about 10 hours so far. Finished around 5:00. On to April!
• Had a short chat with Jay about changes to Ecco and how to submit a pull request through the GitHub website. Maybe I did it right?
• Working on updating experiment code to handle new format – done
• Adding json outputs for ecco data into gpt_experiments
• Got sequence data working:

• Made it a dataframe

         ' pawn'  ' rook'  ' knight'  ' bishop'  ' queen'  ' king'
 knight        1        2          4          3         5        6
 from         36       29          7         13        19       33
 c            10       11          9         13        14       12
4            264     1208        696        865       887      372
 to          668     3314        486        513       325      533
 e            37      160         14         44       102       53
5            944     3567        452       4937      2836     4243
.           1361     3926        933        149      2472     1468
 Black      2512     3164       1508       1604      1974     1925
 moves         7       57         21         49        37       46
 bishop        5        4          2          1         3        6
 from         11        9          7          5        15       23
 b            13       11         12         10        14        9
7           2324     1244        788       1449      2228     1252

GPT Agents:

• Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigm
  • Prevailing methods for mapping large generative language models to supervised tasks may fail to sufficiently probe models’ novel capabilities. Using GPT-3 as a case study, we show that 0-shot prompts can significantly outperform few-shot prompts. We suggest that the function of few-shot examples in these cases is better described as locating an already learned task rather than meta-learning. This analysis motivates rethinking the role of prompts in controlling and evaluating powerful language models. In this work, we discuss methods of prompt programming, emphasizing the usefulness of considering prompts through the lens of natural language. We explore techniques for exploiting the capacity of narratives and cultural anchors to encode nuanced intentions and techniques for encouraging deconstruction of a problem into components before producing a verdict. Informed by this more encompassing theory of prompt programming, we also introduce the idea of a metaprompt that seeds the model to generate its own natural language prompts for a range of tasks. Finally, we discuss how these more general methods of interacting with language models can be incorporated into existing and future benchmarks and practical applications.
• Language models are 0-shot interpreters
  • In this post, I present evidence that the efficacy of 0-shot prompts for GPT-3 has been underestimated, and that more powerful models are more effective at deriving information from 0-shot prompts, while less powerful models have greater need for examples on equivalent tasks. From this evidence, I extrapolate three principal claims:
    • Few-shot prompts are not always an efficient or necessary means of task specification for GPT-3.
    • For some tasks, such as translation between well-known languages, GPT-3 is a 0-shot interpreter – a short task description or signifier suffices to invoke its full capabilities.
    • 0-shot performance scales with model size more drastically than few-shot performance, suggesting that 0-shot task specification will become a more important prompting strategy as language models increase in capability.
• Started training the January 2020 model
• It looks like sim got the new format model trained? It’s up through Feb 21. Need to adjust the query code and parser and do some runs for the queries we discussed last night as well as month/year prompts. And combos, e.g. October 2020 USA [[COVID-19 happened because
• Here’s the first try:

• Clearly, we’re going to have to look across a range of places to get good counts
• Another thought, we could use autoencoding models like BERT to for fill-in-the-blank. Here’s how to do it: How to use BERT from the Hugging Face transformer library

SBIR

• Working on the status report. I’ll distribute tomorrow for input to the financial section

GOES

• 2:00 Meeting
• Still waiting on Vadim to get the reaction wheel efficiency in the right place and inertialess reset

JuryRoom

• Reading “Purakau: Maori Myths Retold by Maori Writers”. Some interesting perspectives on group problem solving and education, particularly in the story ‘Rata’, by Hemi Kelly:
  • ‘“Sail towards the rising sun,” she instructed him, “there you will find Pariroa, the home of Matuku.” After saying this, she handed Rata an old toki, “You will need this to fashion your waka.”’
  • “You didn’t recite the correct karakia – or in fact any karakia. Instead you carelessly chopped down your ancestor, a child of Tāne, for your own gain without offering anything in return.”
  • ‘It’s the same with our rongoā. Anybody can go and pick a leaf and eat it but it’s the process we follow that makes it right. It’s the time we go, the area we visit and the careful selection. The most important thing, though, is our acknowledgement of Tāne through karakia, as it’s the karakia that gives the rongoā its healing properties that make us better.’

GOES

• Register for GSAW – done

SBIR

• More status report

GPT Agents

• Started digging into the GPT-3 documentation. They have a playground which lets you interactively try prompts on the different models. I think this could knowledge could be pulled out in a pretty straightforward way through multiple probes and regex. Here’s some examples:

The great religions of the world are:

Judaism

Christianity

Islam

Hinduism

Buddhism

Sikhism

Jainism

Confucianism

Shinto

A list of the closest religions to Judaism:

Christianity (30%)

Islam (30%)

Buddhism (5%)

Sikhism (5%)

Hinduism (3%)

A list of the closest religions to Christianity:

Judaism

Islam

Hinduism

Buddhism

Agnosticism

Atheism

Christianity

Orthodox

Catholic

Theism

God

• Note that the Judaism and Christianity lists support each other. This could look a lot like the original mapping Java mapping code?
• It does not know about the pandemic (prompt is bold): “coronavirus is a member of the Coronaviridae family, which includes animals and birds as known hosts. The virus is a single-stranded, positive-sense RNA virus with a genome of approximately 30 kb. The genome is organized into three segments: S, M, and L.“
• 3:00 meeting today
  • See if I can train up monthly models
  • Create prompts and evaluate their default
  • Run prompts with Ecco for ranking with our relative terms
  • We’re going to try for the social sensing workshop:
    • The social sensing workshop (started in 2015) is a multidisciplinary meeting place that brings together social scientists and computer scientists, interested in social media analysis, around research that interprets social media as measurement instruments. Social media democratized information production offering an unprecedented view into human habits, customs, culture, stances, and indeed descriptions of physical events that transpire in the world. They also give unprecedented opportunities to spread misinformation, influence opinion, distract from the truth, or advance specific agendas, hidden or overt. The potential of social media to influence populations has brought about an interest in understanding information operations; namely, coordinated efforts on social media meant to alter people’s opinions, emotions, or understanding of events. What are scientific foundations for modeling this new communication, measurement, and influence channel? How to utilize information media signals to better understand social systems, communities, and each other? How to identify and mitigate misuse of this medium? What specifically can one measure or influence, what underlying theoretical framework allows one to do so, and what applications are enabled by the endeavor?  Since measurement and influence operations are well-studied in many physical domains, what can one learn from the physical domain (e.g., from the signal processing literature) to enable novel social media analysis methods? This scope brings about new interdisciplinary research challenges and opportunities at the intersection of communication and sensing, social network analysis, information theory, data mining, natural language processing, artificial intelligence, and social sciences. 

Next year this date will be very exciting!

Replace rear tire!

Shopping! Done

Book

• 2:00 Meeting with Michelle
• Need to add something about Nomads by choice and nomads by circumstance. One is pathfinding, and the other is abandonment/expulsion

SBIR

• Bi-monthly report

GPT Agents

• I just got on the OpenAI GPT-3 beta!
• Train a single month using the new data and the small model
• Set few-shot training by drawing selecting all the tweets that contain the phrase %xxxx% in the corpora and subsample as needed to the desired number of examples. Then generate and store the desired number of results
• Use Ecco to illustrate selected examples?
• Update Ecco and issue a pull request – done!
  • Return the dict if html_output = False
  • html_output = True in args
  • Here’s the non-html version

hello, ecco
2021-02-22 09:02:11.376728: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudart64_110.dll
<LMOutput 'one, two, three, four, five, five, six, six, seven, seven, eight, seven, ten, twelve,' # of lm outputs: 7>
{'tokens': [{'token': 'one', 'token_id': 505, 'type': 'input', 'value': '0.22258912', 'position': 0}, {'token': ',', 'token_id': 11, 'type': 'input', 'value': '0.064860135', 'position': 1}, {'token': ' two', 'token_id': 734, 'type': 'input', 'value': '0.15291078', 'position': 2}, {'token': ',', 'token_id': 11, 'type': 'input', 'value': '0.075403504', 'position': 3}, {'token': ' three', 'token_id': 1115, 'type': 'input', 'value': '0.21873675', 'position': 4}, {'token': ',', 'token_id': 11, 'type': 'input', 'value': '0.04952843', 'position': 5}, {'token': ' four', 'token_id': 1440, 'type': 'input', 'value': '0.17954932', 'position': 6}, {'token': ',', 'token_id': 11, 'type': 'input', 'value': '0.03642196', 'position': 7}, {'token': ' five', 'token_id': 1936, 'type': 'output', 'value': '0', 'position': 8}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 9}, {'token': ' five', 'token_id': 1936, 'type': 'output', 'value': '0', 'position': 10}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 11}, {'token': ' six', 'token_id': 2237, 'type': 'output', 'value': '0', 'position': 12}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 13}, {'token': ' six', 'token_id': 2237, 'type': 'output', 'value': '0', 'position': 14}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 15}, {'token': ' seven', 'token_id': 3598, 'type': 'output', 'value': '0', 'position': 16}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 17}, {'token': ' seven', 'token_id': 3598, 'type': 'output', 'value': '0', 'position': 18}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 19}, {'token': ' eight', 'token_id': 3624, 'type': 'output', 'value': '0', 'position': 20}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 21}, {'token': ' seven', 'token_id': 3598, 'type': 'output', 'value': '0', 'position': 22}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 23}, {'token': ' ten', 'token_id': 3478, 'type': 'output', 'value': '0', 'position': 24}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 25}, {'token': ' twelve', 'token_id': 14104, 'type': 'output', 'value': '0', 'position': 26}, {'token': ',', 'token_id': 11, 'type': 'output', 'value': '0', 'position': 27}], 'attributions': [[0.22258912026882172, 0.06486013531684875, 0.15291078388690948, 0.07540350407361984, 0.21873675286769867, 0.04952843114733696, 0.17954932153224945, 0.03642195835709572], [0.21059739589691162, 0.08506321907043457, 0.10517895966768265, 0.06970299780368805, 0.0833553671836853, 0.06893090158700943, 0.11408285796642303, 0.06513398140668869, 0.19795432686805725], [0.18917866051197052, 0.05591090768575668, 0.11437422782182693, 0.06194028630852699, 0.14940014481544495, 0.05109493061900139, 0.13605113327503204, 0.04323190823197365, 0.16186146438121796, 0.036956313997507095], [0.1881534308195114, 0.06915824860334396, 0.09677033871412277, 0.05285586044192314, 0.07482346147298813, 0.05150040239095688, 0.08954611420631409, 0.052106279879808426, 0.09847243875265121, 0.07199375331401825, 0.15461969375610352], [0.17256230115890503, 0.051994744688272476, 0.09864256531000137, 0.0542512908577919, 0.12412769347429276, 0.048067208379507065, 0.11856795847415924, 0.04466724395751953, 0.115085169672966, 0.03303493186831474, 0.10736697167158127, 0.0316319540143013], [0.1601886749267578, 0.061091333627700806, 0.07975628972053528, 0.04445934668183327, 0.06458994001150131, 0.04427047073841095, 0.06633348762989044, 0.04812125489115715, 0.06941992044448853, 0.04545144364237785, 0.05678795278072357, 0.04917420074343681, 0.2103556990623474], [0.15204396843910217, 0.045879215002059937, 0.08276450634002686, 0.04710154980421066, 0.10755057632923126, 0.04356013983488083, 0.10087976604700089, 0.04368861764669418, 0.1039900854229927, 0.03419572487473488, 0.09188895672559738, 0.018628299236297607, 0.10064685344696045, 0.027181735262274742], [0.15202827751636505, 0.054488517343997955, 0.07370518893003464, 0.03688199818134308, 0.06367962807416916, 0.03571085259318352, 0.06372495740652084, 0.03937176987528801, 0.0647173747420311, 0.04080101102590561, 0.04775563254952431, 0.03407425060868263, 0.08399699628353119, 0.05491986870765686, 0.15414364635944366], [0.14423254132270813, 0.043195948004722595, 0.07906319946050644, 0.04169945418834686, 0.0998333990573883, 0.03888460621237755, 0.09324592351913452, 0.040613241493701935, 0.10006645321846008, 0.03402014449238777, 0.08717798441648483, 0.02316855825483799, 0.06652144342660904, 0.01834581419825554, 0.0651322454214096, 0.024798991158604622], [0.13373175263404846, 0.04726942256093025, 0.06662306189537048, 0.03134704381227493, 0.055253904312849045, 0.029025832191109657, 0.05654153227806091, 0.03190178796648979, 0.05787787586450577, 0.03615624085068703, 0.04139142856001854, 0.03552539646625519, 0.05441423878073692, 0.03426354378461838, 0.05245133861899376, 0.0413157157599926, 0.19490987062454224], [0.1345134824514389, 0.04032375290989876, 0.07328703999519348, 0.03726156800985336, 0.08852741122245789, 0.03443199023604393, 0.08518599718809128, 0.03616797551512718, 0.09120426326990128, 0.031830523163080215, 0.07864660024642944, 0.023800017312169075, 0.05850181728601456, 0.02105688489973545, 0.05589877441525459, 0.015616844408214092, 0.06888125836849213, 0.02486378327012062], [0.12652075290679932, 0.04134056717157364, 0.06051903963088989, 0.02772645838558674, 0.050938140600919724, 0.025345684960484505, 0.05245476961135864, 0.026518667116761208, 0.05662158131599426, 0.030446955934166908, 0.044499896466732025, 0.03298814222216606, 0.05097997933626175, 0.03056175820529461, 0.04033765196800232, 0.02774038165807724, 0.07083491235971451, 0.043943196535110474, 0.15968146920204163], [0.12978246808052063, 0.03859121724963188, 0.06958351284265518, 0.03402606025338173, 0.08373469114303589, 0.030794424936175346, 0.07956207543611526, 0.032011084258556366, 0.08528966456651688, 0.02837086282670498, 0.07092872262001038, 0.02181713469326496, 0.053476281464099884, 0.020540550351142883, 0.05198293551802635, 0.018952928483486176, 0.05511363595724106, 0.013460895046591759, 0.05456216633319855, 0.02741875685751438], [0.1180478110909462, 0.04020005092024803, 0.054879143834114075, 0.02707388810813427, 0.047450218349695206, 0.023739859461784363, 0.05038924515247345, 0.024456234648823738, 0.047715943306684494, 0.02747093327343464, 0.03706439211964607, 0.0308961383998394, 0.04147675260901451, 0.031551189720630646, 0.0328393280506134, 0.027200669050216675, 0.0456111878156662, 0.02625013142824173, 0.05143207311630249, 0.03938102349638939, 0.17487379908561707], [0.1171923577785492, 0.034699417650699615, 0.06537913531064987, 0.03106089122593403, 0.0792543962597847, 0.027469279244542122, 0.0716739073395729, 0.02862895093858242, 0.07993895560503006, 0.025614075362682343, 0.06548977643251419, 0.020967641845345497, 0.05116114020347595, 0.02104913629591465, 0.05117170512676239, 0.021118970587849617, 0.05328201875090599, 0.017415646463632584, 0.0483008548617363, 0.010480904020369053, 0.0580831877887249, 0.020567581057548523], [0.10331424325704575, 0.03044593520462513, 0.0515039786696434, 0.020523859187960625, 0.04673514515161514, 0.01886308752000332, 0.04609629884362221, 0.018619094043970108, 0.04607436805963516, 0.020929865539073944, 0.03714223951101303, 0.02411222830414772, 0.04459238797426224, 0.024786628782749176, 0.033377647399902344, 0.023412106558680534, 0.061772845685482025, 0.022222455590963364, 0.050952207297086716, 0.029886750504374504, 0.05736793577671051, 0.043550651520490646, 0.14371803402900696], [0.11561653017997742, 0.03434549644589424, 0.06183779239654541, 0.029705578461289406, 0.07546865940093994, 0.026097867637872696, 0.06843607127666473, 0.027166275307536125, 0.07738236337900162, 0.024367135018110275, 0.06257404386997223, 0.019834842532873154, 0.04871196299791336, 0.019919248297810555, 0.04681717976927757, 0.020517172291874886, 0.049287013709545135, 0.018004678189754486, 0.0437239371240139, 0.01232482586055994, 0.04339848831295967, 0.008752784691751003, 0.04222355782985687, 0.02348649688065052], [0.10086216032505035, 0.03523547574877739, 0.04474911466240883, 0.02235623449087143, 0.03697269409894943, 0.0194859616458416, 0.04178309068083763, 0.018883418291807175, 0.04281749948859215, 0.020283740013837814, 0.037377286702394485, 0.022277342155575752, 0.03504293039441109, 0.02316695638000965, 0.02779640257358551, 0.02297920361161232, 0.033881187438964844, 0.022866230458021164, 0.03282133489847183, 0.02020248956978321, 0.041888438165187836, 0.025234123691916466, 0.04621865972876549, 0.0418705977499485, 0.18294738233089447], [0.09987115859985352, 0.028910968452692032, 0.053824856877326965, 0.025705233216285706, 0.06502629071474075, 0.02236601710319519, 0.05963760241866112, 0.022843429818749428, 0.06642232835292816, 0.021395988762378693, 0.054698407649993896, 0.018247412517666817, 0.04362887516617775, 0.017933566123247147, 0.04240269213914871, 0.018407726660370827, 0.0463126040995121, 0.017563357949256897, 0.04160071536898613, 0.015998894348740578, 0.03565698862075806, 0.014985213056206703, 0.04064951464533806, 0.0129097243770957, 0.0890578106045723, 0.023942623287439346], [0.0809876024723053, 0.03257425129413605, 0.037600934505462646, 0.02191905491054058, 0.03524129465222359, 0.018587255850434303, 0.04728523641824722, 0.017956718802452087, 0.036090362817049026, 0.018309801816940308, 0.03293214365839958, 0.018760766834020615, 0.03602268546819687, 0.019794179126620293, 0.03185930848121643, 0.020158812403678894, 0.02658083476126194, 0.01986663229763508, 0.02638169191777706, 0.019547127187252045, 0.03760283440351486, 0.02026677317917347, 0.0282907634973526, 0.02577134408056736, 0.0830618366599083, 0.037606652826070786, 0.1689431220293045]]}
C:\Program Files\Python\lib\site-packages\sklearn\decomposition\_nmf.py:1077: ConvergenceWarning: Maximum number of iterations 500 reached. Increase it to improve convergence.
  " improve convergence." % max_iter, ConvergenceWarning)
{'tokens': [{'token': 'one', 'token_id': 505, 'type': 'input', 'position': 0}, {'token': ',', 'token_id': 11, 'type': 'input', 'position': 1}, {'token': ' two', 'token_id': 734, 'type': 'input', 'position': 2}, {'token': ',', 'token_id': 11, 'type': 'input', 'position': 3}, {'token': ' three', 'token_id': 1115, 'type': 'input', 'position': 4}, {'token': ',', 'token_id': 11, 'type': 'input', 'position': 5}, {'token': ' four', 'token_id': 1440, 'type': 'input', 'position': 6}, {'token': ',', 'token_id': 11, 'type': 'input', 'position': 7}, {'token': ' five', 'token_id': 1936, 'type': 'output', 'position': 8}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 9}, {'token': ' five', 'token_id': 1936, 'type': 'output', 'position': 10}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 11}, {'token': ' six', 'token_id': 2237, 'type': 'output', 'position': 12}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 13}, {'token': ' six', 'token_id': 2237, 'type': 'output', 'position': 14}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 15}, {'token': ' seven', 'token_id': 3598, 'type': 'output', 'position': 16}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 17}, {'token': ' seven', 'token_id': 3598, 'type': 'output', 'position': 18}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 19}, {'token': ' eight', 'token_id': 3624, 'type': 'output', 'position': 20}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 21}, {'token': ' seven', 'token_id': 3598, 'type': 'output', 'position': 22}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 23}, {'token': ' ten', 'token_id': 3478, 'type': 'output', 'position': 24}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 25}, {'token': ' twelve', 'token_id': 14104, 'type': 'output', 'position': 26}, {'token': ',', 'token_id': 11, 'type': 'output', 'position': 27}], 'factors': [[[1.1749050617218018, 0.0, 4.174020796199329e-05, 0.0, 0.0, 0.0, 0.0, 1.6842181139509194e-05, 1.6842181139509194e-05, 7.907724648248404e-05, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.00021821660629939288, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 8.583216549595818e-05, 0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0006136916927061975, 0.0, 0.0, 0.03745964169502258, 0.004776454996317625, 0.004776454996317625, 0.08487999439239502, 0.0, 0.15284782648086548, 0.006533266510814428, 0.19210265576839447, 0.0, 0.23522864282131195, 0.0, 0.26809075474739075, 0.0, 0.2898506820201874, 0.0, 0.2553434669971466, 0.0005055475048720837, 0.30294179916381836, 0.005593298468738794, 0.03264965862035751, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0, 1.419406771659851, 0.0, 1.559943437576294, 0.0009293985203839839, 0.0009293985203839839, 1.383615255355835, 0.014789805747568607, 0.9984501600265503, 0.0, 0.707604169845581, 0.0, 0.4685404300689697, 0.0, 0.2941336929798126, 0.0, 0.13417471945285797, 0.009529122151434422, 0.10319297760725021, 0.007372288033366203, 0.0, 0.0, 0.06743992865085602, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.008554508909583092, 0.0, 0.0334680899977684, 0.0, 0.15881435573101044, 0.0, 0.23637165129184723, 2.5085502784349956e-05, 0.27648571133613586, 0.002731950953602791, 0.31169313192367554, 0.0, 0.26685473322868347, 0.0], [0.0, 0.0, 0.0, 0.003048897720873356, 0.0, 0.0, 0.0, 0.11994873732328415, 0.11994873732328415, 0.00020270798995625228, 0.43330657482147217, 0.0153672369197011, 0.6093799471855164, 0.01629449985921383, 1.021621823310852, 0.016422705724835396, 1.1039609909057617, 0.0, 0.6456769704818726, 0.0, 0.3512823283672333, 0.0, 0.09890929609537125, 0.0, 0.017450863495469093, 0.011394587345421314, 0.0, 0.0], [0.0, 0.0, 0.696977436542511, 0.006282013840973377, 0.2083570659160614, 0.0, 0.060352765023708344, 0.0, 0.0, 0.0, 0.0012489393120631576, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0014060646062716842, 0.0015231686411425471, 0.00017680390737950802, 0.004380697384476662, 0.0, 0.0006365026929415762, 0.0, 0.0, 0.0, 0.029077952727675438], [0.0, 0.0, 0.0, 0.0, 0.0, 0.5207799673080444, 0.0, 0.6182433366775513, 0.6182433366775513, 0.00682591088116169, 0.5392492413520813, 0.0012549938401207328, 0.4196029305458069, 0.0, 0.1323639303445816, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.042940836399793625, 0.0, 0.0, 0.0, 0.0986984446644783, 0.0], [0.0, 0.0, 0.0, 0.0015103527111932635, 0.0, 0.0, 0.0, 0.0, 0.0, 0.00701142055913806, 0.0, 0.0, 0.0, 0.022583989426493645, 0.0, 0.007366797886788845, 0.00075104262214154, 0.0, 0.0, 0.0035612466745078564, 0.0, 0.13789713382720947, 0.0, 0.018509462475776672, 0.0, 0.7861870527267456, 0.023602478206157684, 0.8958369493484497], [1.92419270206301e-06, 0.8122134208679199, 0.0042172386310994625, 0.1016063317656517, 0.0, 0.0, 0.0, 0.0, 0.0, 0.004280794877558947, 0.030012501403689384, 0.0, 0.011906291358172894, 0.0, 0.0022146401461213827, 0.0, 0.0, 0.0, 0.0005024807178415358, 0.00017763646610546857, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.021231088787317276, 0.0016654416685923934], [0.0, 0.0, 0.0, 1.7516762018203735, 0.003883163910359144, 0.8880190253257751, 0.017067980021238327, 0.3859412372112274, 0.3859412372112274, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.01619700714945793, 0.0, 0.0, 0.0, 0.0, 0.0, 0.042748916894197464, 0.0, 0.025438468903303146, 0.0, 0.0, 0.0]]]}
{'input_tokens': ['one', ',', ' two', ',', ' three', ',', ' four', ',', ' five', ',', ' five', ',', ' six', ',', ' six', ',', ' seven', ',', ' seven', ',', ' eight', ',', ' seven', ',', ' ten', ',', ' twelve', ','], 'output_tokens': ["'one'", "','", "' two'"], 'rankings': array([[ 655,   12,   16],
       [ 380,   17,   29],
       [1365,    2,  184],
       [ 604,    1,  161],
       [ 797,    1,   66],
       [2141,    1,   98]])}

Huri Whakatau

Jason Edward Lewis is a digital media theorist, poet, and software designer. He is the University Research Chair in Computational Media and the Indigenous Future Imaginary as well as Professor of Computation Arts at Concordia University, Montreal. Born and raised in northern California, Lewis is Hawaiian and Samoan. (Publications)

Aboriginal Territories in Cyberspace is an Aboriginally determined research-creation network whose goal is to ensure Indigenous presence in the web pages, online environments, video games, and virtual worlds that comprise cyberspace.

The Initiative for Indigenous Futures (IIF) is a partnership of universities and community organizations dedicated to developing multiple visions of Indigenous peoples tomorrow in order to better understand where we need to go today.

From Interactions of the ACM: The Humboldt Cup: On narrative, taxonomies, and colonial violence:

• Historian of science Londa Schiebinger [6] offers a compelling account of how the creation of racial and gender hierarchies has permeated the construction of biology and medicine as fields of knowledge. Engaging with the taxonomical system devised by Swedish naturalist Carl von Linnaeus, she points out that traits such as the breasts or the skull were subjected to processes of racialization and sexualization in attempts to produce arguments that would justify the subjugation of femininity and of all racialized peoples. Schiebinger argues that scientists were, in fact, fundamental actors in the colonizing process: In describing, classifying, taxonomizing, and representing this so-called new world, European powers sought to claim ownership over lands, peoples, flora, and fauna. Classifying entire groups of animals based on the presence of breasts was a choice; other characteristics could have been highlighted, such as the presence of hair [6]. White patriarchal domination was thus asserted through notebooks, measuring tools, pens, and paintbrushes just as much as it was through firearms.

Made cookies last night. I blame the weather

Book

• Starting on DARPA Balloon challenge – done!
• Ping Alex Rutherford about diversity injection?
• Delayed meeting with Michelle until Monday at 2:00
• Kurt Braddock is a professor of communications at American University, and the author of Weaponized Words: The Strategic Role of Persuasion in Violent Radicalization and Counter-Radicalization. He explains to Bob how these methods work — and why de-radicalization is a tougher challenge than intervening at an earlier stage. (from On the Media)

GOES

• Harassment training

GPT Agents

• Tweaked the twitter queries.
• Got all the Ecco parts working with the chess DB!

• ML Seminar. Fatima might be interested in time series transformers
• Watched and discussed GPT-3: Language Models are Few-Shot Learners (Paper Explained)

Book

• Finished the Google Doodle. Next is balloon challenge.

GOES

• https://datascience.stackexchange.com/questions/51065/what-is-the-positional-encoding-in-the-transformer-model – Positional Encoding, the terms were a bit vague in some other articles
• kazemnejad.com/blog/transformer_architecture_positional_encoding
• From Tensorflow: tutorials/text/transformer
• http://jalammar.github.io/illustrated-transformer/ – For Attention Heads

SBIR

• Meeting on NN architecture

GPT Agents

• Language Models are few-shot learners (video). Lots of good stuff about how to build probes.
• Was able to get the ecco library to work with my local model!

from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer
from ecco.lm import LM

activations=False
attention=False
hidden_states=True
activations_layer_nums=None

model_str = '../models/chess_model'
tokenizer = AutoTokenizer.from_pretrained(model_str)
model = AutoModelForCausalLM.from_pretrained(model_str, output_hidden_states=hidden_states, output_attentions=attention)

lm_kwargs = {
    'collect_activations_flag': activations,
    'collect_activations_layer_nums': activations_layer_nums}
lm = LM(model, tokenizer, **lm_kwargs)

# Input text
text = "Check."

# Generate 100 tokens to complete the input text.
output = lm.generate(text, generate=100, do_sample=True)

print(output)

Had a nice chat with Antonio about an introduction for the special issue

Today we’re excited to release a big update to the Galaxy visualization, an interactive UMAP plot of graph embeddings of books and articles assigned in the Open Syllabus corpus! (This is using the new v2.5 release of the underlying dataset, which also comes out today.) The Galaxy is an attempt to give a 10,000-meter view of the “co-assignment” patterns in the OS data – basically, which books and articles are assigned together in the same courses. By training node embeddings on the citation graph formed from (syllabus, book/article) edges, we can get really high-quality representations of books and articles that capture the ways in which professional instructors use them in the classroom – the types of courses they’re assigned in, the other books they’re paired with, etc.

Working with Antonio on an introduction to the journal

GPT Agents

• Was going to work on adding local models to Ecco, but got derailed
• GPT-3: Language Models are Few-Shot Learners (Paper Explained)
• Downloaded Sim’s new model
• Uploaded/committed the zipped backup
• Run Sim’s Queries and add to spreadsheet (need to create a view that uses text from extended tweet if available):

#to get week number (in range 0-53 ) from date

select count(*) as COUNT, extract(WEEK from  created_at) as WEEK from twitter_root where text like “%chinavirus%” group by extract(WEEK from  created_at) order by extract(WEEK from  created_at);

#to drill down from month to week and week to day for March and April since they have significantly larger volume of data

select count(*) as COUNT, date(created_at) as DATE from twitter_root  where text like “%chinavirus%” and (month(created_at) = 3 or month(created_at) = 4) and year(created_at) = 2020 group by date(created_at) order by date(created_at);

Figured out how to do this:

create or replace  view long_text_view as
select tr.row_id, tr.created_at,
    case when extended_tweet_row_id <> 0 then et.full_text else tr.text end as long_text
    from twitter_root tr
        left join extended_tweet et on tr.extended_tweet_row_id = et.row_id;

This is kind of cool. The chinavirus is most peaky but low volume. The sars-cov-2 is low volume but flatter, and the more common terms are pretty similar with gentler peaks and slower falloff. I want to know what that dip isaround week nine, and the peak at week 5 in the chinavirus plot.

GOES

• Time to really dig into transformers
• http://peterbloem.nl/blog/transformers – for converting a sequence to sequence into a classifier. The videos are very nice!
• https://datascience.stackexchange.com/questions/51065/what-is-the-positional-encoding-in-the-transformer-model – Positional Encoding, the terms were a bit vague in some other articles
• http://jalammar.github.io/illustrated-transformer/ – For Attention Heads

SBIR

• 10:00 Meeting

Local pharmacies online vaccine scheduling:

• CVS: https://www.cvs.com/vaccine/intake/store/cvd-schedule?icid=coronavirus-lp-vaccine-sd-statetool
• Giant: https://reportsonline.queue-it.net/?c=reportsonline&e=giantfoodcovid19&t=https%3A%2F%2Fgiantfoodsched.rxtouch.com%2Frbssched%2Fprogram%2Fcovid19%2FPatient%2FAdvisory&cid=en-US
• Rite Aid: https://www.riteaid.com/pharmacy/apt-scheduler
• Safeway: https://mhealthappointments.com/covidappt
• Walgreens: https://www.walgreens.com/findcare/vaccination/covid-19/location-screening
• Walmart: https://www.walmart.com/cp/1228302

Mass vaccine clinics have been added by the State of Maryland. You can register at: https://massvax.maryland.gov

Book

• Google Doodle and some other diversity injection? I’d love to fine a Maker example

GPT Agents

• Create training data
• 3:30 Meeting. Nice progress. This is pretty cool:

These haven’t been normalized, but there seems to be a big spike for chinavirus early in the pandemic, which is what I’d expect for stampede-like behavior. Coronavirus has more mentions, and it’s also more spread out and slowly starts to die out as it is incorporated into daily life and other things grab our attention. Sars-Cov-2 follows a different pattern of periodic relative higher and lower interest but doesn’t seem to change that much which is what I would expect for explorer behavior.

GOES

• 11:00 Meeting with Vadim. I’d like to work out the best way to adjust the reaction wheel efficiency. Currently it’s done by adjusting angular velocity in AngleController but I think that the efficiency should be adjusted in the sim code. Also, we need to have the “reset” code that can move the vehicle back to the start position without using physics.
• Multi dimensional inputs:
  • Pytorch (stackoverflow)
  • Tensorflow embeddings as an approach?

Nice little number sequence of a day. For the US at least.

Cool thing! This is pageviews.toolforge.org, that lets you look at and compare Wikipedia page views. These are subjects of recent Google Doodles, which shows the power of diversity injection using ‘random’ prompts.

Book

• 2:00 Meeting with Michelle

GOES

• Add code to AngleController for RW efficiency – done. Not sure if the scripted reaction wheel efficiency failures is being applied correctly. Right now it’s in AngleController as a commanded value to the Rwheels. I think it should probably be handled in the simulation proper
• Add padding to DictionaryEntries for variables that are added later – done

GPT Agents

• Still Ingesting
• Intrinsic vs. Extrinsic Structure
• Graphs and networks
• A Comprehensive Survey of Graph Embedding: Problems, Techniques and Applications
• A graph is hypothetical structure that does not exist in the real world. Any real world data representation may be reduced to a graph, but it is NOT a graph. A network topology can be first reduced to a graph and then algorithms like Dijkstra’s and Kruskal’s can be applied to it for various purposes like routing. All real world situations are subsets of graph problems (of that domain). Most problems of networking, especially those concerning routing reduce to a graph based problem. You will never find a graph in real world scenario. But you will find its subsets or derivatives in practical applications.

Jury Room

• Read Jarod Grovers’ COMPX593 Thesis proposal

ML Group

• 3:30 Meeting. Added some topics to the paper list

Diversity injection works!

Named Tensor Notation

Cleora is a general-purpose model for efficient, scalable learning of stable and inductive entity embeddings for heterogeneous relational data.

Book – shoveled snow instead

GPT Agents

• Couldn’t connect with Jay via email? Sent links by Twitter DM
• Got started on ingesting data

GOES

• Modify Maneuver Creator, Writer, and Reader to handle DataDictionary entries
• Had to update ValueAxis to handle reaction wheel efficiency permutations
• Added writing and reading DictionaryEntrys to the DataDictionary class. Pretty happy with it, though I need to work with the history some?

• 2:00 Meeting

Helping out another older PhD student with Sage Advice. Hopefully it’s helpful!

Book

• More sentences and paragraphs

GPT Agents

• Download and ingest latest data (and keep track of row ids!)
• Get TF, HF, Ecco, and drivers working right

GOES

• 3:00 Meeting
• Start adding rw efficiency and other options to the script. I should be able to make dictionary entries directly and set them
• Add rw efficiency multiplier to AngleController. May have to add a method to see if an entry exists?

MDS

• 10:00 Meeting

JuryRoom

• 5:00 Meeting
• Put a list of journals for publications in 2021
• Journal paper is submitted to Elsevier Journal of Cultural Heritage
• Conference paper is submitted to ACM Computer Human Interaction
• Maori SW dev article
  • https://www.strakertranslations.com/