INDEX

Explanations

patterns related to code or markup syntax

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

mathematical and scientific notation

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

+#+#

-0.54

 ModelExpression

-0.46

 defaultstate

-0.46

 betweenstory

-0.45

sizeCache

-0.41

 insuffisamment

-0.41

諏訪

-0.38

Cyfarwyddwr

-0.37

 NSCoder

-0.37

onAttach

-0.37

POSITIVE LOGITS

:✨

0.61

enties

0.52

 varandra

0.52

 tør

0.51

 caña

0.50

 descart

0.49

entieth

0.49

 abordagem

0.49

 desnuda

0.48

 argint

0.47

Activations Density 0.005%

patterns related to code or markup syntax

mathematical and scientific notation

No Comments

No Known Activations