dorsal/arxiv
View SchemaSearching for memories, Sudoku, implicit check-bits, and the iterative use of not-always-correct rapid neural computation
| Authors | J. J. Hopfield |
|---|---|
| Categories | |
| ArXiv ID | q-bio/0609006 |
| URL | https://arxiv.org/abs/q-bio/0609006 |
Abstract
The algorithms that simple feedback neural circuits representing a brain area can rapidly carry out are often adequate to solve only easy problems, and for more difficult problems can return incorrect answers. A new excitatory-inhibitory circuit model of associative memory displays the common human problem of failing to rapidly find a memory when only a small clue is present. The memory model and a related computational network for solving Sudoku puzzles produce answers that contain implicit check-bits in the representation of information across neurons, allowing a rapid evaluation of whether the putative answer is correct or incorrect through a computation related to visual 'pop-out'. This fact may account for our strong psychological feeling of right or wrong when we retrieve a nominal memory from a minimal clue. This information allows more difficult computations or memory retrievals to be done in a serial fashion by using the fast but limited capabilities of a computational module multiple times. The mathematics of the excitatory-inhibitory circuits for associative memory and for Sudoku, both of which are understood in terms of 'energy' or Lyapunov functions, is described in detail.
{
"annotation_id": "c9df0db1-fdc4-4493-8f68-421e766773d7",
"date_created": "2026-03-02T18:01:34.680000Z",
"date_modified": "2026-03-02T18:01:34.680000Z",
"file_hash": "4eae9af1f75c8028cb57f075ccd4d0f23e8c4dab524e7ecb8c828a8b5f50584a",
"private": false,
"record": {
"abstract": "The algorithms that simple feedback neural circuits representing a brain area\ncan rapidly carry out are often adequate to solve only easy problems, and for\nmore difficult problems can return incorrect answers. A new\nexcitatory-inhibitory circuit model of associative memory displays the common\nhuman problem of failing to rapidly find a memory when only a small clue is\npresent. The memory model and a related computational network for solving\nSudoku puzzles produce answers that contain implicit check-bits in the\nrepresentation of information across neurons, allowing a rapid evaluation of\nwhether the putative answer is correct or incorrect through a computation\nrelated to visual \u0027pop-out\u0027. This fact may account for our strong psychological\nfeeling of right or wrong when we retrieve a nominal memory from a minimal\nclue. This information allows more difficult computations or memory retrievals\nto be done in a serial fashion by using the fast but limited capabilities of a\ncomputational module multiple times. The mathematics of the\nexcitatory-inhibitory circuits for associative memory and for Sudoku, both of\nwhich are understood in terms of \u0027energy\u0027 or Lyapunov functions, is described\nin detail.",
"arxiv_id": "q-bio/0609006",
"authors": [
"J. J. Hopfield"
],
"categories": [
"q-bio.NC"
],
"title": "Searching for memories, Sudoku, implicit check-bits, and the iterative use of not-always-correct rapid neural computation",
"url": "https://arxiv.org/abs/q-bio/0609006"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "80f151af-0055-4a1b-8a0f-996b5b14dfc7",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}