INDEX

Explanations

elements related to detailed survey results and data reporting

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-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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ensburg

-0.39

Sor

-0.33

Bronnen

-0.33

isOpen

-0.33

Sig

-0.33

ItemBackground

-0.32

 ASAP

-0.32

Syl

-0.32

Kal

-0.31

sor

-0.31

POSITIVE LOGITS

 vždy

0.75

 zawsze

0.73

 תמיד

0.68

 jeweils

0.68

每次

0.67

 Siempre

0.67

 siempre

0.66

 always

0.64

 sempre

0.63

always

0.63

Activations Density 0.837%

elements related to detailed survey results and data reporting

No Comments

No Known Activations