INDEX

Explanations

references to a specific location, Flushing

oai_token-act-pair · gpt-3.5-turbo

words related to pushing or urging actions or movements

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

lihood

-0.74

heid

-0.68

 rele

-0.68

edom

-0.66

 vari

-0.66

 Survive

-0.62

 Meet

-0.62

 cogn

-0.61

 surv

-0.61

eering

-0.60

POSITIVE LOGITS

ushing

1.18

ushes

1.06

 Meadows

0.97

USH

0.91

ush

0.90

ushed

0.89

vier

0.82

usher

0.78

aukee

0.76

ãĤ¦ãĤ¹

0.76

Activations Density 0.008%

references to a specific location, Flushing

words related to pushing or urging actions or movements

No Comments

No Known Activations