INDEX

Explanations

references to user-directed features and interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 препратки

-0.92

 समीक्षक

-0.86

Personensuche

-0.80

脚注の使い方

-0.77

Personendaten

-0.75

 OFDb

-0.75

PerformLayout

-0.72

Datuak

-0.72

 للاسماء

-0.72

 Wikimedijinoj

-0.71

POSITIVE LOGITS

 stud

0.55

 đỡ

0.54

stdbool

0.50

allows

0.50

pagestyle

0.49

hagy

0.48

ppsala

0.47

stava

0.47

 General

0.47

hésite

0.46

Activations Density 0.273%

references to user-directed features and interactions

No Comments

No Known Activations