INDEX

Explanations

patterns indicating mathematical or algebraic expressions

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

hexadecimal numbers and code

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

 snippetHide

-0.65

 resourceCulture

-0.59

IsMutable

-0.57

 estekak

-0.56

WithIOException

-0.52

AddTagHelper

-0.52

 Normdatei

-0.50

 restTemplate

-0.47

StoreMessageInfo

-0.47

 amen

-0.44

POSITIVE LOGITS

Personendaten

0.53

 thiệu

0.45

findpost

0.43

DataSnapshot

0.41

ens

0.39

Bei

0.37

くると

0.36

Bid

0.36

Consulting

0.36

 Consulting

0.35

Activations Density 0.104%

patterns indicating mathematical or algebraic expressions

hexadecimal numbers and code

No Comments

No Known Activations