INDEX

Explanations

references to the term "sus."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

GTCX

-0.46

'/',

-0.39

 ruangan

-0.39

 Gemü

-0.39

racht

-0.39

dificio

-0.38

 menger

-0.37

 jadx

-0.37

 intptr

-0.36

 ditemp

-0.36

POSITIVE LOGITS

Sus

2.38

Sus

2.28

sus

2.22

sus

1.99

SUS

1.88

 suspect

1.83

SUS

1.72

 suspects

1.64

 susceptibility

1.38

 suspicious

1.34

Activations Density 0.062%

references to the term "sus."

No Comments

No Known Activations