INDEX

Explanations

instances of repeated patterns or filler words in text

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

possessive pronoun followed by noun

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

uxxxx

-0.59

 CreateTagHelper

-0.57

IntoConstraints

-0.57

ValueStyle

-0.56

 ویکی‌آمباردا

-0.56

 nonUne

-0.56

WithIOException

-0.54

BibitemShut

-0.52

IUrlHelper

-0.52

tph

-0.51

POSITIVE LOGITS

 itself

0.37

クリート

0.34

 Erwachsenen

0.34

 olduğunu

0.33

esistenza

0.32

 šķ

0.32

 varmt

0.32

 democrá

0.32

 enfermos

0.31

 invloed

0.31

Activations Density 0.013%

instances of repeated patterns or filler words in text

possessive pronoun followed by noun

No Comments

No Known Activations