INDEX

Explanations

phrases related to health conditions or medical information

oai_token-act-pair · gpt-3.5-turbo

phrases involving people categorized by their conditions or characteristics

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

press

-0.68

abre

-0.64

 roundup

-0.62

review

-0.62

println

-0.61

obi

-0.61

advertising

-0.60

Fed

-0.59

waters

-0.59

POSITIVE LOGITS

stood

1.58

 regard

1.21

drawn

1.14

 whom

1.13

 regards

1.12

standing

1.01

 impunity

1.00

 disabilities

0.97

 respect

0.95

held

0.91

Activations Density 0.123%

phrases related to health conditions or medical information

phrases involving people categorized by their conditions or characteristics

No Comments

No Known Activations