INDEX

Explanations

references to evidence and testimony 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 @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

 للاسماء

-0.53

WebElementEntity

-0.42

 direcciones

-0.40

NameInMap

-0.40

 FINE

-0.39

 AssemblyCulture

-0.39

 besök

-0.39

 fine

-0.38

 väli

-0.38

ViewFeatures

-0.38

POSITIVE LOGITS

 crimes

0.58

 murders

0.58

 homicide

0.58

 murdered

0.57

 murder

0.57

 killings

0.57

 homic

0.56

 crime

0.56

murder

0.54

Crime

0.51

Activations Density 0.325%

references to evidence and testimony in legal contexts

No Comments

No Known Activations