INDEX

Explanations

qualifiers or modifiers suggesting additional information or comparisons

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

出版年

-0.48

 Chriftian

-0.47

 Cæsar

-0.46

 pleaſure

-0.43

 autorytatywna

-0.41

 Anſ

-0.41

 reaſon

-0.40

⟭

-0.39

 leſs

-0.39

 незавершена

-0.39

POSITIVE LOGITS

PostInfinity

0.52

Rüyada

0.52

UrlResolution

0.52

RegressionTest

0.51

zzleHttp

0.50

DIPSETTING

0.49

archiviato

0.48

Obrázky

0.48

ThroughAttribute

0.47

MarshalTo

0.47

Activations Density 0.042%

qualifiers or modifiers suggesting additional information or comparisons

No Comments

No Known Activations