INDEX

Explanations

events or performances

oai_token-act-pair · gpt-3.5-turbo

references to events or performances described as "shows."

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

etheless

-0.90

srf

-0.79

tymology

-0.75

theless

-0.71

SOURCE

-0.68

 hacked

-0.65

 Klux

-0.65

 captcha

-0.64

 embodied

-0.63

iotics

-0.63

POSITIVE LOGITS

biz

1.42

manship

1.34

runners

1.15

alter

1.05

rooms

1.02

runner

1.02

boat

1.01

room

0.98

downs

0.97

grounds

0.91

Activations Density 0.044%

events or performances

references to events or performances described as "shows."

No Comments

No Known Activations