INDEX

Explanations

instances of the word "walk" in various forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

 walk

-3.06

walk

-2.72

Walk

-2.66

 Walk

-2.64

 WALK

-2.45

 walks

-2.44

 walked

-2.31

WALK

-2.30

walks

-2.27

 walking

-2.16

POSITIVE LOGITS

TagMode

0.61

 Ansch

0.46

__*/

0.46

ɳ

0.45

StoreMessageInfo

0.45

횟

0.44

ọt

0.44

 chấm

0.44

gaged

0.43

 \%$.

0.43

Activations Density 0.148%

instances of the word "walk" in various forms

No Comments

No Known Activations