INDEX

Explanations

terminology and references related to civil rights and discrimination laws

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

discrimination against protected groups

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

varargin

-0.32

 CONCER

-0.31

Permission

-0.31

way

-0.29

Noten

-0.28

Größe

-0.28

 Szab

-0.27

님

-0.27

Rö

-0.27

uña

-0.27

POSITIVE LOGITS

 Prejudice

0.64

 homophobic

0.63

 Discrimination

0.60

discrimination

0.59

 odio

0.59

judice

0.58

OCCURRED

0.57

hate

0.54

enschappelijke

0.54

 discrimination

0.54

Activations Density 0.271%

terminology and references related to civil rights and discrimination laws

discrimination against protected groups

No Known Activations