INDEX

Explanations

numeric values, particularly in the context of historical dates and statistical data

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

9th decade, 2000s

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



-0.59

出版年

-0.55

NameValuePair

-0.50

numerusform

-0.49

FieldBuilder

-0.49

发表于

-0.48

parsedMessage

-0.48

yarnpkg

-0.47

tinyos

-0.47

󠁴

-0.46

POSITIVE LOGITS

 cromado

0.47

 Polda

0.42

 Banten

0.42

 navideño

0.42

 coroa

0.41

 pandémie

0.40

adaptiveStyles

0.39

lıkla

0.39

 plásticos

0.38

 pandemia

0.38

Activations Density 0.027%

numeric values, particularly in the context of historical dates and statistical data

9th decade, 2000s

No Comments

No Known Activations