INDEX

Explanations

references to television and radio programming, particularly focusing on awards and soap operas

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

ValueStyle

-0.48

autorité

-0.43

makeConstraints

-0.39

zysta

-0.38

 xấu

-0.38

 aikaa

-0.36

 насељу

-0.36

 ilman

-0.35

 autorytatywna

-0.35

UrlResolution

-0.35

POSITIVE LOGITS

+#+

0.52

 teleno

0.51

 télé

0.46

scribers

0.46

MLLoader

0.45

 وتسجيلات

0.44

onatos

0.44

 userSchema

0.44

WebServlet

0.43

Familie

0.43

Activations Density 0.003%

references to television and radio programming, particularly focusing on awards and soap operas

No Comments

No Known Activations