INDEX

Explanations

words related to projects or plans

oai_token-act-pair · gpt-3.5-turbo

sentence-ending punctuation marks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 diseng

-0.73

 belly

-0.70

 aven

-0.67

 corrosion

-0.67

umed

-0.66

amus

-0.65

kee

-0.65

ihil

-0.64

 screwed

-0.64

wip

-0.64

POSITIVE LOGITS

 Including

1.50

 These

1.19

 Notably

1.19

 Examples

1.18

 Among

1.16

 Each

1.13

 Particularly

1.07

 Along

1.06

 Those

1.05

 Whether

1.03

Activations Density 0.673%

words related to projects or plans

sentence-ending punctuation marks

No Comments

No Known Activations