INDEX

Explanations

programming-related terms and definitions

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pochod

-0.31

cluye

-0.30

 vengan

-0.29

 Todes

-0.28

 setempat

-0.28

@"

-0.28

by

-0.28

 Económica

-0.28

 meurt

-0.27

POSITIVE LOGITS

 AssemblyTitle

1.29

rungsseite

1.12

Personendaten

1.05

httphttps

1.02

AndEndTag

0.98

msgTypes

0.95

 AssemblyProduct

0.91

richTextPanel

0.91

 ब्रेकडाउन

0.89

RTLR

0.89

Activations Density 0.023%

programming-related terms and definitions

No Comments

No Known Activations