INDEX

Explanations

terms related to medical care and healthcare facilities

oai_token-act-pair · gpt-3.5-turbo

references to medical topics and terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

*/(

-0.84

 Mavericks

-0.68

 guiActiveUnfocused

-0.67

Domin

-0.67

lihood

-0.67

footed

-0.65

 borough

-0.64

 Borough

-0.64

lists

-0.63

 Hornets

-0.63

POSITIVE LOGITS

Medic

1.08

inally

1.02

onom

0.98

ilit

0.90

arie

0.87

aceutical

0.87

 medic

0.86

aret

0.85

 Medic

0.83

inals

0.82

Activations Density 0.012%

terms related to medical care and healthcare facilities

references to medical topics and terms

No Comments

No Known Activations