INDEX

Explanations

technical terms related to data output and control

oai_token-act-pair · gpt-3.5-turbo

terms related to outputs, data, and logistics in various contexts

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

Not in Any Lists

No Comments

Negative Logits

sw

-0.71

é¾įå

-0.66

èĢ

-0.65

åľ

-0.65

ski

-0.65

usercontent

-0.61

cl

-0.61

tis

-0.60

SHIP

-0.60

Otherwise

-0.60

POSITIVE LOGITS

ometimes

1.28

omething

1.24

hift

1.18

heet

1.15

pring

1.09

everal

1.05

pace

1.03

afety

1.02

creen

1.00

are

0.99

Activations Density 0.269%

technical terms related to data output and control

terms related to outputs, data, and logistics in various contexts

No Comments

No Known Activations