INDEX

Explanations

elements related to websites and documentation

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

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_32k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.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

 RequestMethod

-0.72

OGND

-0.68

 contextLoads

-0.64

ftagPool

-0.63

 متعلقه

-0.63

ัติ

-0.63

ResumeLayout

-0.63

μισ

-0.61

 médicaux

-0.61

limsy

-0.59

POSITIVE LOGITS

WireFormatLite

0.64

 alone

0.58

 would

0.55

 will

0.53

 createState

0.52

 này

0.51

typeorm

0.50

netbeans

0.50

 actually

0.49

 clearly

0.48

Activations Density 0.457%

elements related to websites and documentation

No Comments

No Known Activations