“While Large Language Models (LLMs) can exhibit impressive proficiency in isolated,
short-term tasks, they often fail to maintain coherent performance over longer time horizons.”
Oh so they’re basically emulating the typical American?
https://arxiv.org/pdf/2502.15840
#ai #tech #deluluideas #arxiv
Oberst Enzian
npub1rt…lvxvw
2025-05-26 18:45:32 UTC
Author Public Key
npub1rt0n384n9z6wl598tmkdemhkz0cy874xdwkjyzh4erw3ahssx2rsdlvxvwPublished at
2025-05-26 18:45:32 UTCEvent JSON
{
"id": "aff2d30a8ce9dfe5738fc20240eb95f6d98cfe79bdc047aa766090e8725b468a",
"pubkey": "1adf389eb328b4efd0a75eecdceef613f043faa66bad220af5c8dd1ede103287",
"created_at": 1748285132,
"kind": 1,
"tags": [
[
"t",
"ai"
],
[
"t",
"tech"
],
[
"t",
"deluluideas"
],
[
"t",
"arxiv"
],
[
"proxy",
"https://mastodon.social/users/oberstenzian/statuses/114575614413489889",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.mostr.pub"
]
],
"content": "“While Large Language Models (LLMs) can exhibit impressive proficiency in isolated,\nshort-term tasks, they often fail to maintain coherent performance over longer time horizons.”\n\nOh so they’re basically emulating the typical American? \n\nhttps://arxiv.org/pdf/2502.15840\n\n#ai #tech #deluluideas #arxiv",
"sig": "0834baa65dff51e2e944ddadbd1209588f8da6671904deffc74f5ca70b5b18389b189e002b40bd489d545ff8f110c7e0a8ccae68997a6bf2eb108bbb74bacd79"
}