INDEX

Explanations

references to criminal activity and justice-related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

elemField

-0.76

èges

-0.73

SpringBootTest

-0.71

شهاد

-0.70

 triom

-0.69

VIAF

-0.69

AddTagHelper

-0.68

országban

-0.68

IBOutlet

-0.67

phrag

-0.67

POSITIVE LOGITS

 criminal

2.24

 Criminal

2.02

criminal

1.94

 CRIMINAL

1.87

Criminal

1.85

 criminals

1.62

 Crimin

1.59

 crimin

1.57

 Crim

1.48

 kriminal

1.41

Activations Density 0.103%

references to criminal activity and justice-related terminology

No Comments

No Known Activations