INDEX

Explanations

terms related to medical patients and their care

oai_token-act-pair · gpt-3.5-turbo

references to patients in various 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

-0.87

oras

-0.84

ģ«

-0.77

dor

-0.77

auga

-0.73

argon

-0.70

å°Ĩ

-0.70

reme

-0.69

amy

-0.69

laws

-0.69

POSITIVE LOGITS

 Patient

0.85

 patient

0.82

killer

0.80

 confidentiality

0.78

 patients

0.77

 Patients

0.73

 psychiatric

0.70

ysis

0.69

bay

0.67

 medication

0.66

Activations Density 0.023%

terms related to medical patients and their care

references to patients in various contexts

No Comments

No Known Activations