INDEX

Explanations

questions or inquiries related to methods and processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 Vance

-0.79

 raft

-0.77

herz

-0.71

 Aires

-0.69

solete

-0.66

 Stein

-0.66

 Schulte

-0.66

CID

-0.65

 Milne

-0.65

 Coolidge

-0.65

POSITIVE LOGITS

how

2.55

HOW

2.01

How

1.88

How

1.84

HOW

1.80

how

1.76

 cómo

1.34

howto

1.33

יצד

1.23

hoe

1.20

Activations Density 0.111%

questions or inquiries related to methods and processes

No Comments

No Known Activations