INDEX

Explanations

references to various groups of people or communities

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

Jereo

-0.77

Obrázky

-0.74

parsedMessage

-0.71

 kasarigan

-0.69

$")

-0.67

 defStyle

-0.66

migrationBuilder

-0.65

oflavin

-0.64

 '\\;'

-0.63

 Roskov

-0.63

POSITIVE LOGITS

who

1.62

 whom

1.20

who

1.04

 którzy

1.01

 kteří

0.97

 ktorí

0.94

 whose

0.85

whom

0.83

 الذين

0.76

 living

0.72

Activations Density 0.855%

references to various groups of people or communities

No Comments

No Known Activations