INDEX

Explanations

positive and negative emotional descriptors

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

descriptive qualities

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 Reſ

-0.69

 purpoſe

-0.68

OGND

-0.63

 ainfi

-0.63

 whoſe

-0.59

TagMode

-0.59

 ſon

-0.58

 Inſ

-0.57

 ſta

-0.56

 Anſ

-0.56

POSITIVE LOGITS

 delightful

1.78

 delightfully

1.24

 charming

1.03

Delight

0.95

 pleasant

0.95

delight

0.86

charming

0.85

 delectable

0.84

 lovely

0.82

 Delight

0.80

Activations Density 0.003%

positive and negative emotional descriptors

descriptive qualities

No Comments

No Known Activations