INDEX

Explanations

phrases that indicate similarity or comparison

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

Sentences using comparison words

np_acts-logits-general · gemini-2.0-flash

drawing parallels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

migrationBuilder

-0.43

comics

-0.42

Boi

-0.40

 comics

-0.39

 Ters

-0.39

jsxFileName

-0.38

épendance

-0.36

 osteoarthritis

-0.36

 dioxide

-0.36

 lati

-0.36

POSITIVE LOGITS

 same

0.70

 similarly

0.68

Similarly

0.66

Likewise

0.65

 Similarly

0.65

+#+#

0.63

 Likewise

0.63

 Ebenso

0.59

Same

0.59

same

0.59

Activations Density 0.460%

phrases that indicate similarity or comparison

Sentences using comparison words

drawing parallels

No Comments

No Known Activations