INDEX

Explanations

sequences of numbers or numerical representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

the

-0.58

 adipis

-0.49

omenclature

-0.49

 assist

-0.48

 Utilizamos

-0.48

 pelúcia

-0.46

.~(\

-0.45

 INVENTION

-0.45

 analysis

-0.45

</tr>

-0.44

POSITIVE LOGITS

1.22

０

0.63

 Zero

0.58

Zero

0.58

 ZERO

0.57

 zero

0.57

 cero

0.56

০

0.56

zero

0.54

zeros

0.53

Activations Density 1.016%

sequences of numbers or numerical representations

No Comments

No Known Activations