INDEX

Explanations

categories or classifications of items that can be compared or distinguished

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

ang

-0.58

kol

-0.54

мен

-0.52

cosa

-0.52

ANG

-0.51

 some

-0.51

écnicas

-0.51

zens

-0.50

Many

-0.50

anskje

-0.50

POSITIVE LOGITS

UrlResolution

0.84

 الحره

0.83

EDEFAULT

0.78

Попис

0.77

-------

0.76

 unknownFields

0.76

Datuak

0.75

)_/¯

0.74

ddots

0.74

WriteTagHelper

0.73

Activations Density 0.131%

categories or classifications of items that can be compared or distinguished

No Comments

No Known Activations