INDEX

Explanations

boolean values indicating presence or absence of features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

周泽

-0.66

"]=

-0.60

",{

-0.59

"][

-0.58

`,

-0.57

╩

-0.57

`/

-0.56

"](

-0.56

"},

-0.56

>",

-0.55

POSITIVE LOGITS

GenerationType

0.49

tagHelperRunner

0.48

printStackTrace

0.46

 whiteColor

0.42

BeginInit

0.40

 repente

0.40

stddef

0.39

 conmigo

0.39

beginPath

0.38

 Tinggi

0.38

Activations Density 0.378%

boolean values indicating presence or absence of features

No Comments

No Known Activations