INDEX

Explanations

words and phrases that indicate specific assumptions or points of reference in a discussion

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

such a

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

蚪

-0.38

StoreMessageInfo

-0.38

RectangleBorder

-0.37

setVerticalGroup

-0.36

TagMode

-0.35

DockStyle

-0.35

Sto

-0.35

Sto

-0.34

Mid

-0.33

zat

-0.31

POSITIVE LOGITS

 للمعارف

0.63

ᅠ

0.58

المناصب

0.55

PointerException

0.54

 Мексичка

0.51

 takimi

0.47

 autorytatywna

0.47

writeFieldEnd

0.47

SharedDtor

0.47

aarrggbb

0.46

Activations Density 0.065%

words and phrases that indicate specific assumptions or points of reference in a discussion

such a

No Comments

No Known Activations