INDEX

Explanations

terms related to medical conditions such as seizures and epilepsy

oai_token-act-pair · gpt-3.5-turbo

terminology related to seizures and epilepsy

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

conn

-0.80

iquette

-0.72

 Polit

-0.70

thren

-0.67

nis

-0.67

ciation

-0.66

ican

-0.66

 Pour

-0.65

estones

-0.64

tips

-0.64

POSITIVE LOGITS

 seizures

1.19

 epilepsy

1.09

 seizure

1.00

 epile

0.98

 lobe

0.75

 disorders

0.75

ota

0.71

unia

0.69

 impaired

0.69

 Canaver

0.68

Activations Density 0.018%

terms related to medical conditions such as seizures and epilepsy

terminology related to seizures and epilepsy

No Comments

No Known Activations