INDEX

Explanations

phrases indicating presence and existence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

enumi

-0.33

ďaka

-0.32

 cerrada

-0.31

 testigo

-0.31

iconque

-0.28

 aloj

-0.28

 desem

-0.28

 descu

-0.28

Anyone

-0.27

Someone

-0.27

POSITIVE LOGITS

RegressionTest

0.75

ſſung

0.69

 ་་

0.66

endphp

0.66

JspWriter

0.64

 tiguan

0.64

transQ

0.64

 nakalista

0.63

 ſind

0.63

 iconLine

0.62

Activations Density 1.332%

phrases indicating presence and existence

No Comments

No Known Activations