INDEX

Explanations

terms related to favorites or preferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

modelBuilder

-0.73

Autoritní

-0.63

-0.61

 للاسماء

-0.61

 ſind

-0.60

 ISNI

-0.60

 hulls

-0.59

 hull

-0.57

PRNewswire

-0.56

umask

-0.56

POSITIVE LOGITS

favorite

1.00

 favorite

1.00

Favorite

0.98

 Favorite

0.96

 favourite

0.95

 FAVORITE

0.90

favourite

0.90

 Favourite

0.90

Favourite

0.85

 preferred

0.82

Activations Density 0.091%

terms related to favorites or preferences

No Comments

No Known Activations