INDEX

Explanations

keywords related to improvement or enhancement

oai_token-act-pair · gpt-3.5-turbo

references to improving or enhancing objects or experiences

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

orsi

-0.81

staking

-0.70

urus

-0.68

lihood

-0.68

pling

-0.68

gling

-0.68

-0.67

 Nanto

-0.66

ontent

-0.65

pta

-0.65

POSITIVE LOGITS

 disappear

1.22

 easier

1.20

 happen

1.19

 shine

1.14

 behave

1.08

 accessible

1.07

 safer

1.07

 clearer

1.05

 smarter

1.05

 usable

1.04

Activations Density 0.115%

keywords related to improvement or enhancement

references to improving or enhancing objects or experiences

No Known Activations