INDEX

Explanations

statements or quotations in text

oai_token-act-pair · gpt-3.5-turbo

dialogue or statements made by individuals

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Vaugh

-0.70

ãĥ¼ãĥĨ

-0.63

Redd

-0.62

amac

-0.60

 condem

-0.60

egu

-0.60

 disadvant

-0.58

Mobil

-0.57

 advoc

-0.57

 lapt

-0.57

POSITIVE LOGITS

 âĢº

0.55

 guiActive

0.49

â

0.49

 âľ

0.48

 crochet

0.46

ye

0.46

 ages

0.45

vernment

0.45

 Â·

0.45

TRI

0.45

Activations Density 0.452%

statements or quotations in text

dialogue or statements made by individuals

No Comments

No Known Activations