INDEX

Explanations

mentions of medical-related phrases, especially focusing on medical marijuana and medical care

oai_token-act-pair · gpt-3.5-turbo

references to medical topics or terms

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

xual

-0.93

Publisher

-0.71

uble

-0.71

 Modes

-0.65

rer

-0.64

 Downing

-0.64

cair

-0.63

 Guth

-0.63

ween

-0.63

hof

-0.62

POSITIVE LOGITS

 marijuana

1.16

 practition

1.08

 examiner

1.08

 Marijuana

1.04

 cannabis

0.94

ysis

0.92

 diagnoses

0.91

 diagnosis

0.91

mal

0.90

 care

0.89

Activations Density 0.038%

mentions of medical-related phrases, especially focusing on medical marijuana and medical care

references to medical topics or terms

No Comments

No Known Activations