INDEX

Explanations

phrases indicating exceptions or contrasting cases in statements

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

all followed by noun

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

 kaarangay

-0.56

 للمعارف

-0.43

contentLoaded

-0.41

providedIn

-0.39

Drive

-0.38

loop

-0.37

already

-0.37

 вже

-0.36

 merak

-0.36

drive

-0.35

POSITIVE LOGITS

 Normdatei

0.58

enumii

0.57

 يتيمه

0.54

aarrggbb

0.53

ModelAdmin

0.48

AndEndTag

0.47

 كومونز

0.47

0.46

AnchorStyles

0.45

WriteTagHelper

0.44

Activations Density 0.035%

phrases indicating exceptions or contrasting cases in statements

all followed by noun

No Comments

No Known Activations