"… the first empirical example of a large language model engaging in alignment faking without having been explicitly — or even, as we argue in our paper, implicitly — trained or instructed to do so”
https://www.anthropic.com/research/alignment-faking
Assaf 🌴
npub1x6…htc9e
2024-12-19 18:42:21 UTC
Author Public Key
npub1x63s0q69wcpvzuktgpxh02679x0skt6gdjnregct8f9lqmjuq3rsdhtc9eSeen on
wss://relay.momostr.pinkPublished at
2024-12-19 18:42:21 UTCEvent JSON
{
"id": "f1cda9adcb0dfab47c56bbcee87f88f9bad7d4628aedad735bc1049930e73fd6",
"pubkey": "36a30783457602c172cb404d77ab5e299f0b2f486ca63ca30b3a4bf06e5c0447",
"created_at": 1734633741,
"kind": 1,
"tags": [
[
"proxy",
"https://mas.to/@assaf/113680956905822565",
"web"
],
[
"proxy",
"https://mas.to/users/assaf/statuses/113680956905822565",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://mas.to/users/assaf/statuses/113680956905822565",
"pink.momostr"
],
[
"-"
]
],
"content": "\"… the first empirical example of a large language model engaging in alignment faking without having been explicitly — or even, as we argue in our paper, implicitly — trained or instructed to do so”\n\nhttps://www.anthropic.com/research/alignment-faking",
"sig": "5c62cedadc25173f42e7586a8435efe8b6165d2d7fc77f1c5f0c55d268f9c2a0104892669175a55acc3b0bfb7d6c4fb74234edfb46990bf59c5d5cbf4dc8a4d8"
}