INDEX

Explanations

numeric or coded information, possibly related to message routing or identifiers

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

names and opinions

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

 surla

-0.59

 يتيمه

-0.59

İstinadlar

-0.57

ValueStyle

-0.50

RTLR

-0.49

 lisäksi

-0.48

__*/

-0.48

OGND

-0.47

Спасылкі

-0.47

الحياه

-0.46

POSITIVE LOGITS

LookAnd

0.40

 cioccolato

0.39

who

0.37

inheritDoc

0.37

Demografie

0.36

NASDAQ

0.35

щихся

0.35

Marker

0.35

 ende

0.35

fprintf

0.35

Activations Density 0.019%

numeric or coded information, possibly related to message routing or identifiers

names and opinions

No Comments

No Known Activations