INDEX

Explanations

phrases that indicate uncertainty or inquiry

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

asking why

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ValueStyle

-0.57

 CreateTagHelper

-0.54

ArgsConstructor

-0.42

letoe

-0.42

WireFormat

-0.40

msgSender

-0.40

 domen

-0.39

domain

-0.39

izr

-0.38

WaitGroup

-0.38

POSITIVE LOGITS

 mengapa

0.55

 seeming

0.52

 kenapa

0.51

 seemingly

0.50

 dlaczego

0.49

 apparente

0.49

 Warum

0.48

Dlaczego

0.47

why

0.47

styleType

0.45

Activations Density 0.256%

phrases that indicate uncertainty or inquiry

asking why

No Comments

No Known Activations