INDEX

Explanations

references to names, codes, or identifiers in a legal or medical context

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

abbreviations and technical terms

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

RegressionTest

-0.52

itinéraire

-0.50

 صوتيه

-0.47

 FetchType

-0.43

paramref

-0.41

 calendriers

-0.40

 Fruits

-0.39

venu

-0.39

 tubes

-0.39

enzen

-0.39

POSITIVE LOGITS

قایناق‌لار

0.51

rungsseite

0.48

:✨

0.48

 насељу

0.45

webElementXpaths

0.40

 ProtoMessage

0.40

thâu

0.40

Kjelder

0.40

 estekak

0.39

UpInside

0.35

Activations Density 0.004%

references to names, codes, or identifiers in a legal or medical context

abbreviations and technical terms

No Comments

No Known Activations