INDEX

Explanations

references to interactive web content and user guidance

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

"ll", "will", "can" after "you" or "we"

np_acts-logits-general · gemini-2.0-flash

you will find

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

 gloire

-0.48

-0.46

mailto

-0.45

MenuInflater

-0.45

updateOne

-0.43

Geor

-0.43

ushima

-0.42

complexContent

-0.41

 ویکی‌پدیای

-0.41

 Semoga

-0.41

POSITIVE LOGITS

 find

1.51

find

1.46

 finds

1.40

Find

1.22

finds

1.17

 finding

1.15

 Find

1.12

finding

1.11

 FIND

1.07

 Finds

1.05

Activations Density 0.121%

references to interactive web content and user guidance

"ll", "will", "can" after "you" or "we"

you will find

No Comments

No Known Activations