INDEX

Explanations

negative attributes or criticism within a text

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Burr

-0.68

 moder

-0.66

 horr

-0.63

 privately

-0.60

 GDDR

-0.60

 graphene

-0.59

 private

-0.59

 Uganda

-0.59

gar

-0.59

dur

-0.59

POSITIVE LOGITS

division

1.10

conference

1.10

season

1.06

sample

1.05

nine

1.05

seven

1.04

goal

1.04

eligible

1.04

league

1.02

team

1.01

Activations Density 0.091%

negative attributes or criticism within a text

No Comments

No Known Activations