INDEX

Explanations

references to user authentication and connectivity

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

remember, days, preferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ClCompile

-0.43

subplots

-0.39

 שוליים

-0.38

bronn

-0.37

 Dave

-0.36

 daggers

-0.36

Dave

-0.36

axe

-0.36

 مرئيه

-0.36

期刊论文

-0.35

POSITIVE LOGITS

 tartalomajánló

0.75

 незавершена

0.71

 remember

0.66

remember

0.65

 ویکی‌پدی

0.58

Remember

0.56

Diwedd

0.55

CloseOperation

0.54

 MessageBoxIcon

0.54

 pamię

0.53

Activations Density 0.003%

references to user authentication and connectivity

remember, days, preferences

No Comments

No Known Activations