INDEX

Explanations

references to discussions or interactions in community or professional settings

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

near "attend", "talk", or "posted"

np_acts-logits-general · gemini-2.0-flash

attend talks and conferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 незавершена

-0.59

 joaat

-0.50

 banho

-0.48

")"

-0.48

,’’

-0.46

 sabbia

-0.46

 télévis

-0.46

uarts

-0.46

 GenerationType

-0.45

 demeure

-0.45

POSITIVE LOGITS

 GitHub

0.77

 developers

0.77

 evange

0.76

 developer

0.75

 Software

0.75

 intptr

0.74

 Developers

0.73

DockStyle

0.72

 Wikime

0.71

 Developer

0.70

Activations Density 1.262%

references to discussions or interactions in community or professional settings

near "attend", "talk", or "posted"

attend talks and conferences

No Comments

No Known Activations