INDEX

Explanations

references to medical professionals, specifically doctors

oai_token-act-pair · gpt-3.5-turbo

references to medical professionals, specifically doctors

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

trap

-0.70

rake

-0.70

________________________________

-0.69

 Burr

-0.66

van

-0.65

uble

-0.65

yr

-0.65

UME

-0.64

 Uncharted

-0.61

 Â¯

-0.60

POSITIVE LOGITS

 doctors

3.63

Doctors

3.12

 physicians

2.98

 Doctors

2.88

 surgeons

2.74

 doctor

2.49

 psychiatrists

2.33

 clinicians

2.31

 physician

2.15

 veterin

2.13

Activations Density 0.012%

references to medical professionals, specifically doctors

references to medical professionals, specifically doctors

No Comments

No Known Activations