INDEX

Explanations

terms indicating the creation or assembly of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 useState

-0.63

umumkan

-0.60

 Kelurahan

-0.60

 وتسجيلات

-0.58

 latter

-0.56

 Assignee

-0.55

tiously

-0.55

Groetjes

-0.54

Geplaatst

-0.54

Superclass

-0.53

POSITIVE LOGITS

posedge

0.68

 betweenstory

0.67

 estimés

0.65

 otomatig

0.61

примеча

0.57

::_('

0.56

AnchorTagHelper

0.56

 asylum

0.56

 nakalista

0.56

expandindo

0.56

Activations Density 0.280%

terms indicating the creation or assembly of items

No Comments

No Known Activations