INDEX

Explanations

words with the prefix "re-" indicating repetition, reversal, or restoration

oai_token-act-pair · gpt-3.5-turbo

words that contain the substring "re."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

BuyableInstoreAndOnline

-0.76

 Phar

-0.70

 Nicarag

-0.69

 Kern

-0.68

DragonMagazine

-0.68

 à¨

-0.66

gag

-0.65

 Pebble

-0.64

DERR

-0.64

JPM

-0.62

POSITIVE LOGITS

aper

1.03

apers

1.03

porter

1.02

becca

0.98

venue

0.98

pos

0.97

leases

0.97

versible

0.96

actor

0.96

ached

0.93

Activations Density 0.057%

words with the prefix "re-" indicating repetition, reversal, or restoration

words that contain the substring "re."

No Comments

No Known Activations