INDEX

Explanations

actions related to formal processes or legal activities

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

actions followed by articles or objects

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

skap

-0.54

ses

-0.52

complexType

-0.52

存于互联网档案馆

-0.47

selt

-0.45

onna

-0.45

tal

-0.45

 متعلقه

-0.45

LLocation

-0.45

sed

-0.44

POSITIVE LOGITS

 <=",

0.65

PerformLayout

0.62

 tartalomajánló

0.52

ConstraintMaker

0.46

 AssemblyProduct

0.43

HomeAsUpEnabled

0.43

?";

0.43



0.43

tagHelperRunner

0.42

 iconLine

0.40

Activations Density 0.114%

actions related to formal processes or legal activities

actions followed by articles or objects

No Comments

No Known Activations