INDEX

Explanations

medical terms and treatments

oai_token-act-pair · gpt-3.5-turbo

terms related to health and medical conditions

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

undrum

-0.67

bernatorial

-0.65

 Masquerade

-0.64

ulhu

-0.60

eanor

-0.59

 Sons

-0.59

UME

-0.59

mma

-0.58

lished

-0.57

agame

-0.57

POSITIVE LOGITS

bugs

0.69

$.

0.68

 footing

0.64

or

0.62

omever

0.61

like

0.59

 buffs

0.59

loads

0.58

 levels

0.57

 ones

0.57

Activations Density 0.916%

medical terms and treatments

terms related to health and medical conditions

No Comments

No Known Activations