INDEX

Explanations

occurences of the word "four."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

ZL

-0.63

elis

-0.55

stdlib

-0.55

 Estimation

-0.55

ivity

-0.54

genesis

-0.52

emia

-0.52

gentes

-0.52

belline

-0.52

zine

-0.51

POSITIVE LOGITS

 four

1.13

 FOUR

0.98

four

0.97

 Four

0.91

Four

0.88

 quatre

0.83

 quatro

0.81

 dört

0.81

FOUR

0.79

 vier

0.79

Activations Density 0.016%

occurences of the word "four."

No Comments

No Known Activations