INDEX

Explanations

contextual indicators related to location or position

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.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

BeginInit

-0.35

 uitges

-0.32

oucí

-0.32

 Schülerinnen

-0.32

 verzorg

-0.30

darah

-0.30

interopRequire

-0.30

derdag

-0.29

 aquilo

-0.29

erweise

-0.29

POSITIVE LOGITS

case

0.85

regards

0.77

 regards

0.75

consider

0.66

between

0.66

 fact

0.66

 stead

0.65

till

0.65

cedent

0.63

anyway

0.61

Activations Density 0.522%

contextual indicators related to location or position

No Comments

No Known Activations