INDEX

Explanations

pronouns and punctuation marks used in programming and scripting contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

LookAnd

-0.51

<tfoot>

-0.42

 perio

-0.40

ftagPool

-0.39

 AssemblyProduct

-0.36

 domestic

-0.35

ⓧ

-0.35

 nowrap

-0.35

 poveznice

-0.34

POSITIVE LOGITS

java

0.68

UnsafeEnabled

0.62

 Italijani

0.55

msgTypes

0.54

cpp

0.53

swift

0.49

cs

0.47

'))

0.46

scala

0.46

cpp

0.45

Activations Density 0.415%

pronouns and punctuation marks used in programming and scripting contexts

No Comments

No Known Activations