INDEX

Explanations

proper nouns, particularly names and references in a text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 talent

-0.54

 menj

-0.52

 estate

-0.49

Infinity

-0.48

 Talent

-0.47

\{\\

-0.47

 talento

-0.47

jede

-0.45

 culture

-0.45

 Culture

-0.45

POSITIVE LOGITS

 seasonally

0.82

 northward

0.79

جغرافيا

0.78

 globales

0.78

GEBURTSDATUM

0.75

:✨

0.73

 southward

0.72

 BoxFit

0.71

kháu

0.71

 anthropogenic

0.70

Activations Density 0.583%

proper nouns, particularly names and references in a text

No Comments

No Known Activations