INDEX

Explanations

structures or patterns resembling significant data elements

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

formatting separators

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_32/width_131k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.32.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 typelib

-0.47

 surla

-0.42

Mob

-0.34

kjø

-0.34

(||

-0.31

AccessorTable

-0.30

Mob

-0.28

ærer

-0.28

 запи

-0.28

 called

-0.28

POSITIVE LOGITS

ValueStyle

0.72

SequentialGroup

0.67

 FetchType

0.58

 EconPapers

0.55

 مشين

0.53

RTLR

0.52

 MainAxisSize

0.52

AnchorStyles

0.51

RectangleBorder

0.51

 fVar

0.51

Activations Density 0.060%

structures or patterns resembling significant data elements

formatting separators

No Comments

No Known Activations