INDEX

Explanations

phrases related to saving or preservation

oai_token-act-pair · gpt-3.5-turbo

expressions related to saving or conserving resources, particularly in terms of time, money, or life

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

æ©Ł

-0.85

interstitial

-0.83

liga

-0.82

RY

-0.70

ertodd

-0.65

prev

-0.65

KS

-0.65

akening

-0.63

secution

-0.63

ple

-0.62

POSITIVE LOGITS

 lives

1.09

 face

0.94

 Lives

0.93

 souls

0.84

 faces

0.76

 money

0.75

Instance

0.75

Save

0.70

 taxpayers

0.65

selves

0.64

Activations Density 0.048%

phrases related to saving or preservation

expressions related to saving or conserving resources, particularly in terms of time, money, or life

No Comments

No Known Activations