INDEX

Explanations

phrases that specify relationships and references in academic or analytical writing

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

this paper section

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

principalColumn

-0.71

GEBURTSDATUM

-0.68

RectangleBorder

-0.68

tangentMode

-0.66

urably

-0.65

onViewCreated

-0.63

 betweenstory

-0.60

✨:

-0.57

abestanden

-0.57

ContentLoaded

-0.56

POSITIVE LOGITS

पया

0.27

 esternos

0.26

TYPED

0.26

 répondit

0.26

 Dostupné

0.26

 natin

0.25

 }}"></

0.25

 article

0.25

Article

0.25

 RIPRODUZIONE

0.23

Activations Density 0.126%

phrases that specify relationships and references in academic or analytical writing

this paper section

No Comments

No Known Activations