INDEX

Explanations

occurrences of the letter 'a' in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

queles

-0.96

quilo

-0.88

quela

-0.65

ない

-0.52

 affrontare

-0.52

 føl

-0.50

 alcuni

-0.49

 alcune

-0.48

me

-0.48

 również

-0.47

POSITIVE LOGITS

few

0.60

 aVar

0.60

lot

0.59

 priori

0.56

 href

0.55

 posteriori

0.49

OGND

0.49

nap

0.49

aaa

0.48

bit

0.47

Activations Density 0.504%

occurrences of the letter 'a' in various contexts

No Comments

No Known Activations