INDEX

Explanations

legal terminology and references to prohibited actions or offenses

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

unauthorized or illegal actions

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

Not in Any Lists

No Comments

Negative Logits

 /*#__

-0.52

OGND

-0.46

はじめに

-0.44

koning

-0.43

WithIOException

-0.41

IContainer

-0.40

LayoutPanel

-0.40

+#+#

-0.40

Tembelea

-0.40

sizeCache

-0.39

POSITIVE LOGITS

 illegal

0.49

 Illegal

0.48

 unauthorized

0.47

Illegal

0.45

illegal

0.45

AutoScaleMode

0.44

الحياه

0.44

 légal

0.42

 unauthorised

0.41

 Unauthorized

0.40

Activations Density 0.219%

legal terminology and references to prohibited actions or offenses

unauthorized or illegal actions

No Comments

No Known Activations