INDEX

Explanations

expressions of apology or regret

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

tagHelperRunner

-0.58

 surla

-0.57

 Wikiseite

-0.56

ništvo

-0.49

meteries

-0.48

 myſelf

-0.48

Identyfik

-0.47

 الحره

-0.47

 تضيفلها

-0.47

migrationBuilder

-0.46

POSITIVE LOGITS

 BrowserModule

0.43

 sorry

0.42

 battery

0.40

Sorry

0.38

 Sorry

0.37

richTextPanel

0.36

tax

0.36

complexContent

0.36

ỡng

0.35

endpush

0.35

Activations Density 0.017%

expressions of apology or regret

No Comments

No Known Activations