INDEX

Explanations

strong, impactful language related to data, growth, and quality assessments in various contexts

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

not only but also

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

ActionCreators

-0.48

 betweenstory

-0.40

 nonetheless

-0.40

 conservé

-0.37

Pom

-0.37

era

-0.36

atru

-0.35

 regardless

-0.35

PyObject

-0.35

Pom

-0.35

POSITIVE LOGITS

BeginContext

0.64

 autorytatywna

0.60

AutoScaleMode

0.57

XmlAccessorType

0.56

 للمعارف

0.56

 otomatig

0.55

 ligiloj

0.54

Derbyniad

0.54

 Signalez

0.49

findpost

0.49

Activations Density 0.068%

strong, impactful language related to data, growth, and quality assessments in various contexts

not only but also

No Comments

No Known Activations