INDEX

Explanations

technical and analytical language used in research studies

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

Followed by attributes or options

np_acts-logits-general · gemini-2.0-flash

free, available, picks, watch, compute

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Gewerbe

-0.34

 casque

-0.34

出版年

-0.33

 Auge

-0.32

うち

-0.32

่านั้น

-0.32

 protoimpl

-0.32

 bancaria

-0.31

 primaria

-0.31

 primario

-0.30

POSITIVE LOGITS

aarrggbb

0.73

 resourceCulture

0.57

قایناق‌لار

0.53

OCCURRED

0.53

iffance

0.52

WireFormatLite

0.50

Jeografia

0.50

 kasarigan

0.49

ViewImports

0.49

ьаж

0.47

Activations Density 0.419%

technical and analytical language used in research studies

Followed by attributes or options

free, available, picks, watch, compute

No Comments

No Known Activations