INDEX

Explanations

specific quantifiable terms and statistical references in the context of data analysis or abstraction

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

greater surveillance

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

tvguidetime

-0.46

rungsseite

-0.44

 Majefty

-0.42

ReusableCell

-0.41

 calendriers

-0.40

 otomatig

-0.39

 asist

-0.39

 препратки

-0.39

retweeted

-0.38

 eingel

-0.37

POSITIVE LOGITS

 Roskov

0.52

MeasureSpec

0.48

ografija

0.46

 pulumi

0.44

crollView

0.44

ьаж

0.43

sizeCache

0.43

səhifə

0.43

reportWebVitals

0.42

putstr

0.41

Activations Density 0.238%

specific quantifiable terms and statistical references in the context of data analysis or abstraction

greater surveillance

No Comments

No Known Activations