INDEX

Explanations

terms associated with bias, subjectivity, or social issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_23/width_16k/average_l0_75

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

úgó

-0.58

 بيها

-0.58

ôles

-0.54

Lugares

-0.51

volent

-0.51

AZIONI

-0.51

forums

-0.51

Contexte

-0.51

міністра

-0.50

///</

-0.50

POSITIVE LOGITS

iness

0.99

fulness

0.92

liness

0.91

yness

0.90

ishness

0.90

ness

0.89

ousness

0.87

lessness

0.83

veness

0.83

uality

0.83

Activations Density 0.404%

terms associated with bias, subjectivity, or social issues

No Comments

No Known Activations