INDEX

Explanations

references to legal terminology and issues related to reputation and defamation

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

Negative comments or accusations

np_acts-logits-general · gemini-2.0-flash

personal attacks and insults

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 parad

-0.55

Plen

-0.51

原始内容存档于

-0.50

ύπ

-0.46

dbh

-0.45

IntoConstraints

-0.45

 Struct

-0.45

 Negoti

-0.44

liculas

-0.44

Fug

-0.44

POSITIVE LOGITS

 slander

1.27

 insults

1.11

 attacks

1.06

 insulting

1.05

 defamation

1.05

 hate

1.03

 insult

1.01

 hateful

1.01

 derogatory

1.00

 hatred

0.99

Activations Density 0.555%

references to legal terminology and issues related to reputation and defamation

Negative comments or accusations

personal attacks and insults

No Comments

No Known Activations