INDEX

Explanations

terms related to health risks and medical conditions

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

biological and medical research

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

AndEndTag

-0.60

Enllaços

-0.56

 հղումներ

-0.50

帖最后由

-0.50

endregion

-0.48

 betweenstory

-0.47

Спољашње

-0.46

Bakgrunnsstoff

-0.46

 something

-0.45

 Basically

-0.44

POSITIVE LOGITS

 translateY

0.43

然而

0.43

 efficacious

0.42

 valueForKey

0.40

近年来

0.40

 paradig

0.38

 concurrently

0.37

 ſind

0.37

 stakeholders

0.37

 pourtant

0.36

Activations Density 1.018%

terms related to health risks and medical conditions

biological and medical research

No Comments

No Known Activations

terms related to health risks and medical conditions

biological and medical research

No Comments

No Known Activations