INDEX

Explanations

phrases related to drugs or medicine

oai_token-act-pair · gpt-3.5-turbo

mentions of rugs

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

 Highlands

-0.64

 Reconstruction

-0.62

 Eclipse

-0.62

 Dresden

-0.62

 Korean

-0.61

 oral

-0.59

 Petra

-0.59

 Grail

-0.59

cav

-0.59

 Peninsula

-0.59

POSITIVE LOGITS

rug

1.48

uay

0.90

ular

0.87

ulent

0.86

ulence

0.85

iosity

0.85

icz

0.85

iety

0.85

atism

0.83

osity

0.82

Activations Density 0.005%

phrases related to drugs or medicine

mentions of rugs

No Comments

No Known Activations