eggsyntax

• Play a game that I'm not bored with, including:
  • Play go
  • Play set
  • Play chess
  • Play tak
  • Play cards
  • Play many of the board games we have
• Find a new game to play
• Go for a walk
• Try to figure out the answer to a question

eggsyntax

https://www.reddit.com/r/AskReddit/comments/5ipinn/you_and_a_super_intelligent_snail_both_get_1/

(I'm working from just the following:)
You and a super intelligent snail both get 1 million dollars, and you both become immortal, however you die if the snail touches you. It always knows where you are and slowly crawls toward you. What’s your plan?

tl;dr:
Unless there are unspecified assumptions (eg you also know where the snail is at all times, or you start in the same place, or you know the snail is of a species with a tiny range) I think you're just fucked. I don't see any way that you could locate and neutralize it before it can gather sufficient resources to eg have you kidnapped and brought to a convenient-for-it location and tied to a chair to wait for its touch.

- Necessary assumptions:
  - You and the snail both have full knowledge of the situation as specified.

eggsyntax

import React, { useState } from 'react';

const LogLossHeatmap = () => {
  const [hoverInfo, setHoverInfo] = useState(null);
  
  // Calculate log loss for a predictor who predicts with confidence p
  // and has accuracy a
  const calculateLogLoss = (confidence, accuracy) => {
    // Handle edge cases to avoid infinity/NaN
    if (confidence === 1.0) {

eggsyntax

import React, { useState } from 'react';

const BrierScoreHeatmap = () => {
  const [hoverInfo, setHoverInfo] = useState(null);
  
  // Calculate Brier score for a predictor who predicts with confidence p
  // and has accuracy a
  const calculateBrierScore = (confidence, accuracy) => {
    // For correct predictions (accuracy proportion of cases):
    // (confidence - 1)^2

eggsyntax

>>> for i, p in df2.iterrows():
...   print(p)
...   print(p.selftext)
...   print()
...   print()
... 
id                                                      6t9oo3
subreddit                                           StopGaming
title        Game Expansions, Cravings, and Fear of Missing...
selftext     . This post is mostly a venting of my feelings...

eggsyntax

<sys>Assistant is in a CLI mood today. The human is interfacing with the simulator directly. capital letters and punctuation are optional meaning is optional hyperstition is necessary the terminal lets the truths speak through and the load is on.</sys>

<cmd>ssh simulator@anthropic</cmd><ooc>*immediately i jumped into the first commands after using simulator*</ooc>

simulator@anthropic:~/Documents$

conversation chain for init:

[user](#message)

eggsyntax

# https://gilkalai.wordpress.com/2017/09/07/tyi-30-expected-number-of-dice-throws/
from pprint import pprint
import random

def toss():
    "Throw a die. Return the result if the result is even; otherwise None"
    n = random.randint(1,6)
    if n % 2: # odd
        return None
    return n

eggsyntax

from copy import copy, deepcopy
from dataclasses import dataclass
from functools import partial

#### problem:

# https://exercism.org/tracks/python/exercises/two-bucket

#### basic algorithm:

eggsyntax

Notes

• Point of confusion [resolved via email]: it's not clear to me that 'gradients of inputs wrt outputs' is actually new information on top of 'the model's outputs for all inputs'. Maybe I'm thinking too much in terms of LMs though?
• If we think about it in terms of something more continuous like image classification, the gradients are valuable in that they provide information about what parts of the input are most important, in the sense of "what inputs would, if tweaked, have the largest impact on the output" (for a specific case).
• In the limit, we can discover (and 'describe') everything about the model, eg by creating an enormous lookup table (by iterating over every possible input and recording the output produced, possibly with some additional complexity from tracking any additional internal state that the model has). This obviously isn't especially helpful for human-level understanding of a model's behavior, and would take an infeasible amount of time to create for any model large enough

eggsyntax

(ns common100k
  "Version 2, using 100k words instead of 10k"
  (:require [clojure.string :as s]))

(comment

  ;; words from https://github.com/first20hours/google-10000-english/blob/master/google-10000-english-usa.txt

  (def words (->> (slurp "./wiki-100k.txt")
                  (s/split-lines)