INDEX

Explanations

expressions and phrases indicating offense or emotional reactions to statements or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 Auss

-0.45

 rostro

-0.45

不像

-0.43

setTypeface

-0.43

 usually

-0.42

 several

-0.42

 во

-0.42

Usually

-0.41

郎

-0.41

lik

-0.41

POSITIVE LOGITS

GEBURTSDATUM

1.12

 autorytatywna

0.92

MessageTagHelper

0.90

 doInBackground

0.87

 initComponents

0.84

+#+#

0.83

 виправивши

0.80

styleType

0.79

出版年

0.78

StoryboardSegue

0.77

Activations Density 0.231%

expressions and phrases indicating offense or emotional reactions to statements or situations

No Comments

No Known Activations