Skip to content

Instantly share code, notes, and snippets.

View Nikolaj-K's full-sized avatar
💭
I'm combinating Why's.

Nikolaj Kuntner Nikolaj-K

💭
I'm combinating Why's.
119 followers · 15 following
View GitHub Profile
@Nikolaj-K
Nikolaj-K / historic_market_data.py
Last active December 13, 2017 20:14
work with historic crypto market data
'''
git clone
https://github.com/jhogan4288/coinmarketcap-history
python2.7 coinmarketcap_usd_history.py neo 2016-09-09 2017-12-12 > neo.csv
'''
import csv
import matplotlib.pyplot as plt
@Nikolaj-K
Nikolaj-K / coz-contributions-2017_12-11-to-2017_12_17.md
Last active December 19, 2017 21:56
@Nikolaj-K
Nikolaj-K / CoZ-contributions-1-7-2018.md
Created January 8, 2018 21:25
CoZ-contributions-1-7-2018.md
@Nikolaj-K
Nikolaj-K / coz-report-2018-4.md
Last active January 29, 2018 17:20
coz-report-2018-4.md

Presented in Vienna with Chris, Dean & Erik

(that was 2 weeks ago - I forgot to make a gist - but this week we have video proof of the event!)

Tech talks and Q&A at the Vienna meetup - part 1 and 2

Video 1:

https://youtu.be/PMvx-H2qgNQ

(Overview at 3:56, talk starts at 11:04)

@Nikolaj-K
Nikolaj-K / reward_model.py
Last active January 29, 2018 21:41
reward_model.py
import matplotlib.pyplot as plt
def bonusNormalized(levelHalf, level):
return 1 - levelHalf / float(levelHalf + level)
def levelFactor(levelHalf, level, bonusMax):
return 1 + bonusMax * bonusNormalized(levelHalf, level)
def reward(levelHalf, bonusMax, marketFactor, levelOld, levelDelta):
@Nikolaj-K
Nikolaj-K / imusify_reward_model_v2.py
Created February 14, 2018 01:32
imusify reward model 2.0 dummy reference implementation
import math
ICO_SUPPLY = 10**9
DECIMALS = 8
FLOAT_FACTOR = 10**DECIMALS
IMUSIFY_OPERATION_STACK = (3-1) * ICO_SUPPLY * FLOAT_FACTOR
SOME_USER_ADDRESS = 'ASsudg...poor soul'
IMUSIFY_MAIN_IMU_ADDRESS = "AXjsdb...much IMU here"
IMUSIFY_OPERATION_IMU_ADDRESS = "AKjsdb...less IMU here"
@Nikolaj-K
Nikolaj-K / ftm_dag.py
Last active June 9, 2019 01:26
ftm/dag ratio plot with prices from cmc
import matplotlib.pyplot as plt
PRICE_TODAY = dict(dag=[0.003599], ftm=[0.021852])
def parse(name):
filename = "{}.csv".format(name)
prices = [float(line[13:21]) for line in open(filename)][::-1] # line[13:21] cuts out the right column values
return [y / prices[0] for y in prices + PRICE_TODAY[name]]
dag = parse("dag")
@Nikolaj-K
Nikolaj-K / network1.py
Last active May 28, 2023 13:33
A ff neural network in vanilla Python
"""
A compact feedforward AI. Script explained here:
https://youtu.be/z2aq21lMw40
In this video we're implementing a feed forward neutral network that is able to discern
handwritten digits in vanilla Python.
References:
http://static.latexstudio.net/article/2018/0912/neuralnetworksanddeeplearning.pdf # deep learning book
@Nikolaj-K
Nikolaj-K / questions_to_ask_yourself.md
Last active January 12, 2024 13:07
Generic project and viva questions to ask yourself

Generic viva questions

This file contains generic questions about your work, thesis, project, etc. and they are presented in this video:

https://youtu.be/gqTihL-7FbA

To compile it, I split some sensible thesis questions into reasonable categories and then added good stuff from the top ten or so internet search results for "viva questions" on university and news articles. Many questions are very similar, but each with a particular spin to them.

@Nikolaj-K
Nikolaj-K / MIRI_papers.md
Last active October 8, 2022 23:03
Sorted TODO reading list

Original list on the MIRI website:

https://intelligence.org/all-publications/

E Hubinger, C van Merwijk, V Mikulik, J Skalse, and S Garrabrant. 2019. “Risks from Learned Optimization in Advanced Machine Learning Systems.” arXiv:1906.01820 [cs.AI].

39 pages ++

V Kosoy. 2019. “Delegative Reinforcement Learning: Learning to Avoid Traps with a Little Help.” Forthcoming at the Safe Machine Learning workshop at ICLR.