INDEX

Explanations

phrases indicating capability or ability to perform tasks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 userSchema

-0.72

WireFormat

-0.69

 velkommen

-0.60

 Paglinawan

-0.58

 quema

-0.58

Burns

-0.58

бище

-0.57

Kraj

-0.57

 Majefty

-0.57

 Numerology

-0.57

POSITIVE LOGITS

can

1.19

Can

1.15

Can

1.03

 able

1.03

又能

0.93

 ability

0.92

can

0.91

CAN

0.90

ecan

0.89

能

0.89

Activations Density 0.126%

phrases indicating capability or ability to perform tasks

No Comments

No Known Activations