INDEX

Explanations

phrases indicating expectations or outcomes related to user actions

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

Follows "will" or "ll"

np_acts-logits-general · gemini-2.0-flash

will find you'll receive notice

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

weetened

-0.56

tifies

-0.51

InlineData

-0.49

лоси

-0.49

 refuses

-0.48

 telling

-0.47

PreferredItem

-0.46

flink

-0.46

indata

-0.45

Controllo

-0.45

POSITIVE LOGITS

 receive

1.03

 receives

0.91

 received

0.86

 Receive

0.85

 RECEIVE

0.82

receive

0.81

 receiving

0.80

 otrzy

0.77

 Receiving

0.75

Receive

0.73

Activations Density 0.258%

phrases indicating expectations or outcomes related to user actions

Follows "will" or "ll"

will find you'll receive notice

No Comments

No Known Activations