INDEX

Explanations

technical terms and parameters associated with models and equations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 للاسماء

-0.51

pushFollow

-0.46

enders

-0.43

richTextPanel

-0.42

 Roskov

-0.42

 tả

-0.42

ьаж

-0.41

inola

-0.39

 Circ

-0.39

 Trag

-0.38

POSITIVE LOGITS

 coefficient

1.61

 coefficients

1.54

coefficient

1.28

 Coefficients

1.26

 Coefficient

1.22

 coeff

1.22

系数

1.22

coefficients

1.20

 coef

1.16

Coefficient

1.13

Activations Density 2.022%

technical terms and parameters associated with models and equations

No Comments

No Known Activations