INDEX

Explanations

references to invisible abilities and their effects

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

Comes before abilities or strengths descriptions

np_acts-logits-general · gemini-2.0-flash

allows for Kiki

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 createState

-0.76

सन्दर्भ

-0.70

новништво

-0.65

CGRectMake

-0.64

makeConstraints

-0.63

IsContent

-0.63

 Superhost

-0.63

FontStyle

-0.63

WriteBarrier

-0.63

 Aristote

-0.63

POSITIVE LOGITS

Rüyada

0.59

artige

0.59

NOPQRST

0.57

 mergeFrom

0.56

lichkeit

0.54

🏻

0.53

tua

0.52

🏼

0.52

 tartalomajánló

0.52

تفصیلات

0.52

Activations Density 4.272%

references to invisible abilities and their effects

Comes before abilities or strengths descriptions

allows for Kiki

No Comments

No Known Activations