INDEX

Explanations

technical terms and code-related language

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

Code or special characters followed by specific words

np_acts-logits-general · gemini-2.0-flash

in $

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

ConstraintMaker

-0.89

 nakalista

-0.78

 للاسماء

-0.72

 betweenstory

-0.68



-0.68

 كومونز

-0.65

WebServlet

-0.63

 propOrder

-0.62

StoryboardSegue

-0.61

RTLI

-0.60

POSITIVE LOGITS

 chez

0.36

in

0.33

://

0.32

 chaves

0.32

QLabel

0.32

UnsafeEnabled

0.31

enregistrement

0.31

 Contactez

0.31

 muestras

0.31

his

0.30

Activations Density 1.310%

technical terms and code-related language

Code or special characters followed by specific words

in $

No Comments

No Known Activations