INDEX

Explanations

medical terms or references

oai_token-act-pair · gpt-3.5-turbo

references to medical themes or terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Jobs

-0.82

argon

-0.76

===

-0.75

RP

-0.74

 Password

-0.72

yre

-0.70

 Byte

-0.68

 Canter

-0.68

uitous

-0.67

Boo

-0.66

POSITIVE LOGITS

med

4.01

Med

1.98

Med

1.98

 meditation

1.69

med

1.63

 medic

1.47

 medi

1.44

 mediation

1.38

 Meditation

1.30

MED

1.28

Activations Density 0.009%

medical terms or references

references to medical themes or terminology

No Comments

No Known Activations