INDEX

Explanations

references to abuse in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 resourceCulture

-0.75

 ویکی‌پدی

-0.66

WriteAttribute

-0.62

MessageTagHelper

-0.61

 ſans

-0.59

 Autorisations

-0.58

новништво

-0.58

 समीक्षाओं

-0.57

 pinulongan

-0.56

OGND

-0.56

POSITIVE LOGITS

 Jewish

0.58

Jewish

0.57

°

0.50

 jewish

0.48

 abuse

0.48

abuse

0.46

 abusing

0.45

 jüdis

0.44

 leather

0.43

Rubber

0.43

Activations Density 0.182%

references to abuse in various contexts

No Comments

No Known Activations