INDEX

Explanations

references to familial and caretaking relationships

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

Adverbs of frequency

np_acts-logits-general · gemini-2.0-flash

sometimes or occasionally

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

-0.59

writeHead

-0.49

"]();

-0.48

 вас

-0.47

 вам

-0.46

дитесь

-0.46

 tadi

-0.45

 dieux

-0.44

 Your

-0.44

ชี

-0.43

POSITIVE LOGITS

 whenever

1.42

whenever

1.29

 Whenever

1.25

Whenever

1.22

Occasionally

1.10

 Occasionally

1.05

 sometimes

1.05

 occasionally

1.02

sometimes

0.95

Sometimes

0.95

Activations Density 0.442%

references to familial and caretaking relationships

Adverbs of frequency

sometimes or occasionally

No Comments

No Known Activations