INDEX

Explanations

terms related to specific names and entities, such as "CPI," "Maya," "suitcase," and "Mavericks."

oai_token-act-pair · gpt-3.5-turbo

specific references to organizations and notable entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ebook

-1.07

zynski

-0.89

sten

-0.85

nan

-0.83

enic

-0.80

bley

-0.79

culosis

-0.76

ressor

-0.75

idas

-0.75

soDeliveryDate

-0.74

POSITIVE LOGITS

CPI

0.94

 slic

0.75

Fey

0.73

 clicks

0.72

================================

0.71

 Mavericks

0.71

++++++++

0.69

Notable

0.68

Minimum

0.67

Scroll

0.66

Activations Density 0.022%

terms related to specific names and entities, such as "CPI," "Maya," "suitcase," and "Mavericks."

specific references to organizations and notable entities

No Comments

No Known Activations