INDEX

Explanations

Medical terms and references, particularly related to medication or medical professionals

oai_token-act-pair · gpt-3.5-turbo

references to medical professionals or medical themes

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

 sovere

-0.70

agra

-0.67

 infer

-0.65

 vend

-0.63

 redes

-0.63

 warrants

-0.61

 affidav

-0.60

VER

-0.60

âĨĴ

-0.60

istries

-0.60

POSITIVE LOGITS

 Medic

0.85

ports

0.71

mith

0.70

swick

0.69

wolf

0.68

metics

0.68

stadt

0.68

orthy

0.68

enberg

0.68

gamer

0.67

Activations Density 0.000%

Medical terms and references, particularly related to medication or medical professionals

references to medical professionals or medical themes

No Comments

No Known Activations