INDEX

Explanations

quantitative assessments or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 iſt

-1.13

 ſind

-1.07

Efq

-1.07

 ―――――

-1.03

ſelf

-1.03

Portale

-0.99

 itſelf

-0.99

GEBURTSDATUM

-0.97

-0.96

ſelves

-0.96

POSITIVE LOGITS

0.79

0.76

 something

0.72

0.71

....

0.71

0.69

ish

0.69

or

0.69

0.68

Activations Density 0.306%

quantitative assessments or measurements

No Comments

No Known Activations