INDEX

Explanations

mentions of new items, features, or updates

oai_token-act-pair · gpt-3.5-turbo

references to new items, features, or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

aminer

-0.82

Zip

-0.80

ashtra

-0.78

uddin

-0.75

ascript

-0.66

itta

-0.66

ollah

-0.65

acebook

-0.64

ppe

-0.63

 actionGroup

-0.63

POSITIVE LOGITS

bies

1.24

bie

1.17

 arrivals

0.99

 additions

0.97

 Zealand

0.96

foundland

0.96

 entrants

0.88

 discoveries

0.86

 YORK

0.85

ãĤ´

0.80

Activations Density 0.107%

mentions of new items, features, or updates

references to new items, features, or concepts

No Comments

No Known Activations