INDEX

Explanations

entities mentioned alongside a specific numerical value

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate authorship or attribution in historical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

vc

-0.80

Netflix

-0.75

igun

-0.75

ploy

-0.74

verage

-0.72

internet

-0.71

ahime

-0.71

wcs

-0.69

register

-0.68

TPP

-0.68

POSITIVE LOGITS

 Johann

1.38

 Augustus

1.35

 Herod

1.33

 Raphael

1.28

Sir

1.26

 Napoleon

1.25

 Rabbi

1.25

 Moses

1.25

 Franz

1.24

Ibn

1.23

Activations Density 0.370%

entities mentioned alongside a specific numerical value

phrases that indicate authorship or attribution in historical contexts

No Comments

No Known Activations