INDEX

Explanations

numerical intervals

oai_token-act-pair · gpt-3.5-turbo

terms related to intervals and distributions in a statistical context

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

umption

-0.79

 Unic

-0.74

son

-0.74

tty

-0.72

ongyang

-0.72

gling

-0.71

advertising

-0.69

isel

-0.68

ISA

-0.68

NetMessage

-0.68

POSITIVE LOGITS

 intervals

1.45

 interval

1.40

 ende

0.78

 pauses

0.72

 appe

0.72

 mund

0.71

 answ

0.71

 occas

0.71

 timer

0.70

 eleph

0.70

Activations Density 0.009%

numerical intervals

terms related to intervals and distributions in a statistical context

No Comments

No Known Activations