INDEX

Explanations

proper nouns and specific identifiers related to rules and policies

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

PreferredItem

-0.44

 sacré

-0.39

chaik

-0.38

setLayoutManager

-0.38

 moindre

-0.37

Bioaccumulative

-0.37

szönöm

-0.37

Conservative

-0.36

auto

-0.36

 sapin

-0.36

POSITIVE LOGITS

 kaarangay

0.62

 gethan

0.52

 propOrder

0.50

 Italijanski

0.48

ftagPool

0.44

 zelve

0.44

WebElementEntity

0.44

AddHtmlAttribute

0.44

 ligiloj

0.44

PostInfinity

0.43

Activations Density 0.003%

proper nouns and specific identifiers related to rules and policies

No Comments

No Known Activations