INDEX

Explanations

symbols and formatting elements in code or structured data

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

arrows and structure

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

 myſelf

-0.48

spesies

-0.44

 itſelf

-0.42

MutableLiveData

-0.39

 chofe

-0.39

ſy

-0.37

Kapcsolódó

-0.37

extAlignment

-0.34

ſelves

-0.34

 verſ

-0.34

POSITIVE LOGITS

 defaultstate

0.66

 nahilalakip

0.65

EndContext

0.63

󠁣

0.52

autoreleasepool

0.52

errHandler

0.49

ワイイ

0.48

NameInMap

0.48

HostException

0.48

ParallelGroup

0.47

Activations Density 0.125%

symbols and formatting elements in code or structured data

arrows and structure

No Comments

No Known Activations