INDEX

Explanations

programming constructs related to control flow and iteration in code

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

Code-related punctuation and increment operators

np_acts-logits-general · gemini-2.0-flash

programming code and each

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

exitRule

-0.49

 PeEnEo

-0.47

 MainAxisSize

-0.45

DeleteBehavior

-0.45

 Destroyer

-0.43

 Mixer

-0.42

acuzzi

-0.42

 makeStyles

-0.42

RTSN

-0.42

adaptiveStyles

-0.42

POSITIVE LOGITS

++){

0.54

++){

0.52

++)

0.51

AnchorStyles

0.50

forEach

0.49

++)

0.45

 Chaque

0.43

Each

0.42

 each

0.42

each

0.41

Activations Density 0.175%

programming constructs related to control flow and iteration in code

Code-related punctuation and increment operators

programming code and each

No Comments

No Known Activations