INDEX

Explanations

references to mandatory information submission and guidelines

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

finding facts and use

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

est

-0.41

osis

-0.36

ⓘ

-0.35

แค

-0.34

 revolved

-0.34

ecap

-0.33

mapbox

-0.33

 doğ

-0.33

("$.

-0.33

<em>

-0.33

POSITIVE LOGITS

 MainAxisSize

0.68

awtextra

0.68

invokeLater

0.68

GEBURTSDATUM

0.60

 pleaſure

0.59

 Offisielt

0.57

 NSCoder

0.56

 Administrativna

0.56

httphttps

0.55

 CreateTagHelper

0.54

Activations Density 0.022%

references to mandatory information submission and guidelines

finding facts and use

No Comments

No Known Activations