INDEX

Explanations

the beginning of a document or text segment

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

Followed by punctuation or spacing

np_acts-logits-general · gemini-2.0-flash

Japanese rooftop, archaic English

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.57

<eos>

-0.51

core

-0.50

-0.49

 online

-0.49

in

-0.48

 body

-0.48

-0.47

aus

-0.47

POSITIVE LOGITS

 Majefty

1.03

Efq

1.01

setVerticalGroup

0.99

 pleaſure

0.99

 myſelf

0.96

 ―――――

0.94

 ་་

0.93

 Roskov

0.91

ſelf

0.91

$_"

0.90

Activations Density 0.085%

the beginning of a document or text segment

Followed by punctuation or spacing

Japanese rooftop, archaic English

No Comments

No Known Activations