INDEX

Explanations

instances where something is described or categorized in a specific way

oai_token-act-pair · gpt-3.5-turbo

phrases that describe states of being or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

abases

-0.77

CBC

-0.71

istries

-0.71

 Toledo

-0.68

 Carnival

-0.67

cycle

-0.67

 Saskatchewan

-0.66

aukee

-0.63

 Ribbon

-0.62

 Lansing

-0.62

POSITIVE LOGITS

yip

0.66

inventoryQuantity

0.66

 precaution

0.64

 influenced

0.63

 progressed

0.62

 progresses

0.61

 aloud

0.61

acer

0.61

 threatened

0.60

 indicated

0.60

Activations Density 0.095%

instances where something is described or categorized in a specific way

phrases that describe states of being or conditions

No Comments

No Known Activations