INDEX

Explanations

instances of the word "start"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "start."

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

phy

-0.76

illard

-0.74

ugs

-0.65

iliary

-0.63

asca

-0.63

itsch

-0.62

obi

-0.60

 entirety

-0.60

ocally

-0.59

otropic

-0.59

POSITIVE LOGITS

nings

1.21

ners

0.90

starting

0.83

rek

0.78

 anew

0.76

UP

0.75

around

0.75

watch

0.74

ribune

0.72

points

0.70

Activations Density 0.056%

instances of the word "start"

occurrences of the word "start."

No Known Activations