INDEX

Explanations

medical professionals and procedures

oai_token-act-pair · gpt-3.5-turbo

references to medical professionals and their qualifications or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

bum

-0.75

 citiz

-0.71

 Rebellion

-0.71

bitious

-0.70

PLA

-0.70

 momentum

-0.70

izons

-0.70

Grab

-0.69

amaz

-0.69

 Thief

-0.69

POSITIVE LOGITS

 prescribe

1.61

 diagnose

1.60

 advise

1.40

 diagnoses

1.34

 specialize

1.32

 presc

1.27

 prescribing

1.27

 certify

1.24

 prescribed

1.24

 recommend

1.20

Activations Density 0.489%

medical professionals and procedures

references to medical professionals and their qualifications or actions

No Comments

No Known Activations