INDEX

Explanations

technical terms and error messages related to programming or data processing

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

debug, info, test

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

DockStyle

-0.56

 BrowserModule

-0.50

 tilgjenge

-0.48

ască

-0.42

⟬

-0.41

 piele

-0.41

⤹

-0.41

velocityX

-0.40

 dostup

-0.38

umán

-0.37

POSITIVE LOGITS

uxxxx

0.79

Personensuche

0.57

 CanadaChoose

0.57

 tartalomajánló

0.55

ंदीखरीदारी

0.53

 Roskov

0.52

:✨

0.51

InjectAttribute

0.50

 дописавши

0.48

httphttps

0.47

Activations Density 0.006%

technical terms and error messages related to programming or data processing

debug, info, test

No Comments

No Known Activations