INDEX

Explanations

references to news articles or reports

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

(continued...)

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

presentation

-0.44

存于互联网档案馆

-0.40

 presentation

-0.37

presenta

-0.36

resentation

-0.36

 вста

-0.35

tagPool

-0.35

 CommonModule

-0.34

pointments

-0.33

fits

-0.32

POSITIVE LOGITS

 EClass

0.57

 ivelany

0.56

IRQn

0.53

writeFieldEnd

0.50

 отредактировал

0.50

 idéia

0.49

devamını

0.49

Viitteet

0.48

SuspendLayout

0.48

+#+#

0.46

Activations Density 0.002%

references to news articles or reports

(continued...)

No Comments

No Known Activations