INDEX

Explanations

data structure elements, likely related to JSON formatting and nested information

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

direction and location

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

OGND

-0.80

تقاوى

-0.78

astéroïdes

-0.77

aarrggbb

-0.77

 Biôgrafia

-0.76



-0.74

Vidite

-0.72

 Савезне

-0.66

 Italijanski

-0.65

WithIOException

-0.65

POSITIVE LOGITS

 south

0.47

 north

0.41

 South

0.40

 near

0.40

rungsseite

0.40

 jauh

0.39

south

0.39

homonymie

0.38

 vicino

0.37

 nearby

0.35

Activations Density 0.014%

data structure elements, likely related to JSON formatting and nested information

direction and location

No Comments

No Known Activations