INDEX

Explanations

instances of the verb "be" in various forms

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

"be" followed by an adjective

np_acts-logits-general · gemini-2.0-flash

be + descriptor of state

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

 itſelf

-0.97

 ftill

-0.83

 ſtill

-0.78

 fometimes

-0.74

 alſo

-0.73

 themſelves

-0.72

 Jefus

-0.69

 himſelf

-0.69

 fortfarande

-0.69

 myſelf

-0.69

POSITIVE LOGITS

 able

1.33

1.02

 seen

0.92

an

0.91

 considered

0.91

 called

0.84

the

0.83

 made

0.79

 held

0.79

 used

0.78

Activations Density 0.423%

instances of the verb "be" in various forms

"be" followed by an adjective

be + descriptor of state

No Comments

No Known Activations