INDEX

Explanations

dialogue or quotes from individuals in various contexts

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

attribution of speech

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

 المعيارى

-0.71

 שוליים

-0.49

VIAF

-0.47

tides

-0.46

 kapture

-0.45

adaptiveStyles

-0.44

tagHelperRunner

-0.44

 poffe

-0.43

Cyr

-0.43

 esternos

-0.42

POSITIVE LOGITS

setVerticalGroup

0.60

LookAnd

0.45

tagext

0.43

 AssemblyCulture

0.42

 Roskov

0.41

脚注の使い方

0.40

protoimpl

0.40

 ویکی‌پدی

0.38

parsedMessage

0.37

erweise

0.37

Activations Density 0.013%

dialogue or quotes from individuals in various contexts

attribution of speech

No Comments

No Known Activations