INDEX

Explanations

medical terms and mentions of medical professionals

oai_token-act-pair · gpt-3.5-turbo

references to medical topics and health-related terms

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

earchers

-0.96

*/(

-0.82

é¾

-0.75

DAQ

-0.67

ibly

-0.66

soever

-0.66

izont

-0.63

IRO

-0.63

emouth

-0.62

 Wrap

-0.61

POSITIVE LOGITS

onom

0.90

racted

0.81

zzi

0.80

onomic

0.80

ature

0.80

Medic

0.78

otic

0.74

stown

0.70

oto

0.70

Brew

0.69

Activations Density 0.051%

medical terms and mentions of medical professionals

references to medical topics and health-related terms

No Comments

No Known Activations