INDEX

Explanations

sequentially numbered items or sections in text

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

Roman numerals in lists

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

yntaxException

-0.66

 مرئيه

-0.64

 queſta

-0.63

ſicht

-0.62

 BorderSide

-0.60

CloseOperation

-0.59

PMailer

-0.59

[@BOS@]

-0.59

<unused28>

-0.58

ConstraintMaker

-0.58

POSITIVE LOGITS

subsubsection

0.43

 indigenous

0.32

viaf

0.32

impanan

0.31

 terceira

0.29

 replacement

0.29

 Weeks

0.29

 Visser

0.28

 high

0.28

Activations Density 0.423%

sequentially numbered items or sections in text

Roman numerals in lists

No Comments

No Known Activations