INDEX

Explanations

references to specific types of entities or subjects, particularly in an academic or scientific context

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

vermectin

-0.41

 Rigidbody

-0.39

rrggbb

-0.39

 strijd

-0.38

 morti

-0.36

 appunto

-0.35

 oorlog

-0.35

 relationship

-0.35

poptosis

-0.35

 mesmos

-0.35

POSITIVE LOGITS

+#+#

0.71

 tartalomajánló

0.69

hyrchwyd

0.64

ंदीखरीदारी

0.56

 Мексичка

0.55

ientras

0.55

uxxxx

0.55

UnsafeEnabled

0.54

 архивлан

0.54

pingente

0.54

Activations Density 0.406%

references to specific types of entities or subjects, particularly in an academic or scientific context

No Comments

No Known Activations