INDEX

Explanations

categories or identifiers in text formatting

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

digital object identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

Predecesor

-0.43

ⓧ

-0.40

Personensuche

-0.39

<bos>

-0.38

her

-0.38

 iprot

-0.37

 OFDb

-0.37

 Wikimedijinoj

-0.37

sī

-0.37

IUrlHelper

-0.36

POSITIVE LOGITS

 authors

0.54

 صوتيه

0.52

出版年

0.50

WEBPACK

0.49

ThroughAttribute

0.49

authors

0.48

قایناق‌لار

0.48

SourceChecksum

0.47

providedIn

0.46

casus

0.46

Activations Density 0.041%

categories or identifiers in text formatting

digital object identifiers

No Comments

No Known Activations