INDEX

Explanations

proper nouns, especially names of characters and places, frequently associated with dramatic contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ligiloj

-0.50

usuf

-0.48

 onPostExecute

-0.46

PullParser

-0.46

paksa

-0.45

xygen

-0.44

dirond

-0.44

 bezig

-0.43

 Einfluß

-0.43

__*/

-0.43

POSITIVE LOGITS

expandindo

0.58

ami

0.57

 OFDb

0.52

GEBURTSDATUM

0.52

 densities

0.49

でも

0.48

 invoice

0.48

 Worms

0.47

ENS

0.46

 worms

0.46

Activations Density 0.479%

proper nouns, especially names of characters and places, frequently associated with dramatic contexts

No Comments

No Known Activations