INDEX

Explanations

sentences that indicate the presence of a concluding thought or summary within a text

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.12.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

chanind/pile-uncopyrighted-gemma-1024-abbrv-1B

Embeds

IFrame

Link

Negative Logits

 findes

-0.56

piscina

-0.55

 nahilalakip

-0.54

 Suiza

-0.54

Manuscript

-0.52

 fundido

-0.51

IndentedString

-0.51

vábbi

-0.51

 Urdu

-0.51

 urbanos

-0.50

POSITIVE LOGITS

*/].

0.54

__*/

0.52

GetLength

0.52

 consultato

0.50

ursal

0.47

StoryboardSegue

0.46

 NSError

0.46

 adhé

0.44

AppMethodBeat

0.44

्री

0.43

Activations Density 0.540%

sentences that indicate the presence of a concluding thought or summary within a text

No Known Activations