INDEX

Explanations

words associated with names and proper nouns

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

Accented characters

np_acts-logits-general · gemini-2.0-flash

Júlio Delgado

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

iería

-0.48

ómica

-0.47

RegressionTest

-0.47

 many

-0.45

frow

-0.45

 חיצוניים

-0.45

antine

-0.44

 vanden

-0.44

 Green

-0.44

 कारण

-0.42

POSITIVE LOGITS

RenderAtEndOf

0.70

jooq

0.67

homonymie

0.63

存于互联网档案馆

0.63

 ſche

0.61

 ſind

0.61

 Reſ

0.61

ſelf

0.60

WithIOException

0.59

ſelves

0.59

Activations Density 0.102%

words associated with names and proper nouns

Accented characters

Júlio Delgado

No Comments

No Known Activations