INDEX

Explanations

dialogue or conversational elements, particularly related to communication and advice

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

after variants of "tell"

np_acts-logits-general · gemini-2.0-flash

informing or telling

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

ConverterFactory

-0.57

 оригіналу

-0.56

Datuak

-0.56

ViewFeatures

-0.55

 खरीदा

-0.49

 calendriers

-0.47

setVerticalGroup

-0.46

 krav

-0.44

cketed

-0.43

 cré

-0.43

POSITIVE LOGITS

 tell

2.80

 telling

2.55

 Tell

2.46

Tell

2.45

 told

2.44

 tells

2.38

 TELL

2.36

 Telling

2.30

Telling

2.22

tell

2.21

Activations Density 0.552%

dialogue or conversational elements, particularly related to communication and advice

after variants of "tell"

informing or telling

No Comments

No Known Activations