INDEX

Explanations

actions related to stepping or taking initiative

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

Movement-related words following "step"

np_acts-logits-general · gemini-2.0-flash

step into, out, back, up

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

 nakalista

-0.84

GEBURTSDATUM

-0.80

WithIOException

-0.74

 jScrollPane

-0.73

AndEndTag

-0.69

rrggbb

-0.67

XmlAccessorType

-0.67

MigrationBuilder

-0.67

 Italijanski

-0.65

енча

-0.65

POSITIVE LOGITS

 into

0.82

 oneself

0.60

 foot

0.56

out

0.54

 myself

0.53

off

0.53

 back

0.50

 piede

0.50

 ourselves

0.49

up

0.49

Activations Density 0.143%

actions related to stepping or taking initiative

Movement-related words following "step"

step into, out, back, up

No Comments

No Known Activations