INDEX

Explanations

programming and coding constructs related to function definitions and return statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

[];

-0.54

 OMITBAD

-0.54

=[];

-0.53

]]

-0.51

esModule

-0.50

"}>

-0.50

 betweenstory

-0.50

}.

-0.50

VYMaps

-0.49

;%

-0.48

POSITIVE LOGITS

->

0.51

offsetof

0.43

printf

0.40

 NULL

0.35

tinyos

0.33

請繼續往下閱讀

0.33

sizeof

0.32

 diensten

0.32

 sizeof

0.32

ULL

0.32

Activations Density 0.620%

programming and coding constructs related to function definitions and return statements

No Comments

No Known Activations