INDEX

Explanations

references to "principal" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 wireType

-0.54

SharedDtor

-0.54

makeText

-0.49

 poesía

-0.49

 poète

-0.48

télé

-0.46

 Deixe

-0.46

 Déf

-0.46

ientras

-0.45

 Egipto

-0.45

POSITIVE LOGITS

 principal

2.94

 Principal

2.86

Principal

2.84

principal

2.73

 PRINCIPAL

2.58

 Principals

2.11

 principals

2.02

PRINCIP

1.93

Princip

1.68

princip

1.66

Activations Density 0.115%

references to "principal" in various contexts

No Comments

No Known Activations