INDEX

Explanations

phrases related to architectural structures or physical entities

oai_token-act-pair · gpt-3.5-turbo

references to domes and related structural terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

packing

-0.79

âĢł

-0.71

supp

-0.70

VD

-0.70

Quotes

-0.70

TRY

-0.70

ters

-0.67

UGH

-0.67

ACTED

-0.64

Cly

-0.64

POSITIVE LOGITS

dom

1.10

ially

0.89

loo

0.82

iaries

0.80

doms

0.76

ingen

0.76

dyn

0.75

entric

0.74

 domin

0.72

bub

0.71

Activations Density 0.008%

phrases related to architectural structures or physical entities

references to domes and related structural terms

No Comments

No Known Activations