INDEX

Explanations

expressions of commitment and loyalty

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

 ModelExpression

-0.67

 שוליים

-0.64

Попис

-0.62

majánló

-0.61

Хьажоргаш

-0.60

Personensuche

-0.58

 InputDecoration

-0.57

 Monfieur

-0.57

 noDo

-0.56

wiliwch

-0.55

POSITIVE LOGITS

 fortsatt

0.41

 commitment

0.40

 intends

0.39

 compromisso

0.38

 willing

0.38

 willingness

0.38

 committed

0.37

 following

0.37

 compromiso

0.37

 dispuestos

0.36

Activations Density 0.011%

expressions of commitment and loyalty

No Comments

No Known Activations