INDEX

Explanations

references to morning news programming

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

otoro

-0.85

躇

-0.84

.*")]

-0.81

 Croft

-0.80

########.

-0.79

HasForeignKey

-0.79

CrossRef

-0.79

chengladbach

-0.78

 politiet

-0.78

enumi

-0.78

POSITIVE LOGITS

 morning

2.12

 Morning

1.93

 MORNING

1.81

morning

1.80

 mornings

1.78

Morning

1.76

 Morn

1.34

 Morgen

1.29

 morn

1.26

 mattina

1.21

Activations Density 0.037%

references to morning news programming

No Comments

No Known Activations