INDEX

Explanations

references to seized evidence and narcotics in legal contexts

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

found or seized evidence

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

WriteAttribute

-0.45

 betweenstory

-0.45

medriver

-0.44

mobileqq

-0.44

 Poitiers

-0.44

 Caterina

-0.42

 HTMLElement

-0.42

 '\\;'

-0.42

GEBURTSDATUM

-0.41

CloseOperation

-0.41

POSITIVE LOGITS

 confiscated

0.74

 seized

0.58

 confis

0.56

 confiscation

0.51

 contraband

0.41

 Seiz

0.41

ioutil

0.40

 disambiguazione

0.39

 Seizures

0.37

 seizing

0.37

Activations Density 0.026%

references to seized evidence and narcotics in legal contexts

found or seized evidence

No Comments

No Known Activations