INDEX

Explanations

sections of text that represent data or numerical formats

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

configuration assignment statements

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

ScopeManager

-0.62

enderror

-0.57

 ***!

-0.56

complexContent

-0.53

ImageContext

-0.50

IsContent

-0.47

 AssemblyCulture

-0.46

StackNavigator

-0.44

aarrggbb

-0.43

лтамалар

-0.42

POSITIVE LOGITS

antası

0.45

 étoient

0.38

ulemon

0.38

tangentMode

0.38

InjectAttribute

0.37

 Nachmittag

0.35

utilisons

0.35

 étrangère

0.34

annique

0.34

BeforeMethod

0.33

Activations Density 0.001%

sections of text that represent data or numerical formats

configuration assignment statements

No Comments

No Known Activations