INDEX

Explanations

words related to encryption, ransomware, and specific product names

oai_token-act-pair · gpt-3.5-turbo

terms related to various types of medical conditions or phenomena, particularly those associated with 'otic' conditions

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

creen

-0.72

 WARN

-0.71

rising

-0.70

neapolis

-0.69

HUD

-0.68

 largeDownload

-0.67

edin

-0.67

Rate

-0.65

fman

-0.65

ĵĺ

-0.65

POSITIVE LOGITS

otic

1.29

otics

1.06

ism

0.72

OPLE

0.70

osis

0.69

ishly

0.68

 Goods

0.67

otine

0.67

ional

0.66

astically

0.65

Activations Density 0.012%

words related to encryption, ransomware, and specific product names

terms related to various types of medical conditions or phenomena, particularly those associated with 'otic' conditions

No Comments

No Known Activations