INDEX

Explanations

occurrences of the letters 'sm' followed by a number, with higher activations for 'sm' followed by 9 and 10

oai_token-act-pair · gpt-3.5-turbo

references to smoky or hazardous conditions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.81

quickShipAvailable

-0.73

EMENT

-0.73

 Emir

-0.69

OPLE

-0.67

docker

-0.67

 Scrib

-0.67

ktop

-0.66

HCR

-0.65

WER

-0.64

POSITIVE LOGITS

oky

1.08

sm

1.04

itten

0.91

attering

0.88

elly

0.85

udge

0.84

ithing

0.83

ixtures

0.83

ench

0.82

ugg

0.82

Activations Density 0.005%

occurrences of the letters 'sm' followed by a number, with higher activations for 'sm' followed by 9 and 10

references to smoky or hazardous conditions

No Comments

No Known Activations