INDEX

Explanations

expressions of emotional support and relationships

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

followed by a comma or conjunction

np_acts-logits-general · gemini-2.0-flash

feeling or state of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 stammen

-0.49

ızlık

-0.46

 erfolgen

-0.46

 folly

-0.45

hols

-0.42

stdlib

-0.42

 gelingt

-0.41

referer

-0.41

includes

-0.40

cresc

-0.40

POSITIVE LOGITS

 excited

1.57

 afraid

1.56

 aware

1.56

 interested

1.50

 scared

1.44

 unable

1.44

 bored

1.42

 willing

1.41

 intrigued

1.39

 concerned

1.39

Activations Density 0.358%

expressions of emotional support and relationships

followed by a comma or conjunction

feeling or state of being

No Comments

No Known Activations