INDEX

Explanations

references to legal concepts and contexts

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

legal-year-location citations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

Expressions

-0.38

ruptedException

-0.35

 friction

-0.34

res

-0.34

 control

-0.34

Erreferentziak

-0.34

 pega

-0.34

 पुन

-0.33

uen

-0.33

AntiForgeryToken

-0.33

POSITIVE LOGITS



0.68

Autoritní

0.68

 autorytatywna

0.60

GEBURTSDATUM

0.58

GTCX

0.58

GOTREF

0.57

bootstrapcdn

0.56

AndEndTag

0.53

afficheront

0.53

StoreMessageInfo

0.51

Activations Density 0.009%

references to legal concepts and contexts

legal-year-location citations

No Comments

No Known Activations