INDEX

Explanations

terms related to medical patients, treatments, and healthcare

oai_token-act-pair · gpt-3.5-turbo

references to patients in medical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

angular

-0.73

ģ«

-0.69

GGGG

-0.68

ãĥ¼ãĥĨãĤ£

-0.68

SourceFile

-0.68

sg

-0.66

shell

-0.66

dust

-0.62

OPA

-0.60

pol

-0.60

POSITIVE LOGITS

 patients

0.97

 diagnosed

0.90

ients

0.89

 Patients

0.86

 hospitalized

0.81

ysis

0.80

iology

0.78

 patient

0.74

ele

0.74

ubs

0.74

Activations Density 0.024%

terms related to medical patients, treatments, and healthcare

references to patients in medical contexts

No Comments

No Known Activations