INDEX

Explanations

phrases indicating relationships or affiliations among subjects

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

 poitrine

-0.53

 Sturm

-0.53

 kering

-0.51

 MIPS

-0.50

 cốc

-0.49

 فريبيس

-0.48

colt

-0.47

 prostu

-0.47

 noce

-0.46

Enders

-0.46

POSITIVE LOGITS

setVerticalGroup

0.77

のうち

0.71

HostException

0.70

 MessageBoxIcon

0.69

/−

0.68

 INSEE

0.67

 közül

0.65

うち

0.63

當中

0.62

 становника

0.62

Activations Density 0.375%

phrases indicating relationships or affiliations among subjects

No Comments

No Known Activations