INDEX

Explanations

names of people and locations in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 resultMap

-0.73

HAT

-0.70

 Bremen

-0.70

Rango

-0.69

YI

-0.69

Frat

-0.69

ntry

-0.69

træ

-0.68

 Trin

-0.68

 Intru

-0.68

POSITIVE LOGITS

 وتسجيلات

0.84

Crus

0.83

UnusedPrivate

0.79

crus

0.74

 sores

0.73

—

0.73

 nakalista

0.72

 phosphoric

0.72

 Crusaders

0.70

 către

0.70

Activations Density 1.395%

names of people and locations in context

No Comments

No Known Activations