INDEX

Explanations

mandatory language or words indicating obligation in legal contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

Vikt

-0.73

Tril

-0.66

Lina

-0.66

Spinning

-0.65

ubio

-0.63

Olivia

-0.62

Spin

-0.62

Idio

-0.61

 Diagonal

-0.60

Donnie

-0.60

POSITIVE LOGITS

 shall

0.59

detectChanges

0.57

 Shall

0.50

 CURIAM

0.48

__.__

0.47

0.46

 defStyleAttr

0.46

retudo

0.46

permitAll

0.45

}());

0.44

Activations Density 0.476%

mandatory language or words indicating obligation in legal contexts

No Comments

No Known Activations