INDEX

Explanations

specific named entities, such as names of people or places

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

internal

-0.87

Interstitial

-0.77

essler

-0.77

mbuds

-0.76

soDeliveryDate

-0.71

usa

-0.71

Newsletter

-0.70

idae

-0.70

thood

-0.68

Zup

-0.68

POSITIVE LOGITS

 actor

1.15

 musician

1.15

 youngster

1.14

 singer

1.13

 Frenchman

1.13

 outspoken

1.11

 former

1.11

 rapper

1.10

 actress

1.06

 comedian

1.06

Activations Density 0.409%

specific named entities, such as names of people or places

No Comments

No Known Activations