INDEX

Explanations

phrases that indicate a strong or noticeable difference

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing sharp contrasts or significant differences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

hops

-0.85

sembly

-0.74

imester

-0.67

 diligently

-0.66

llular

-0.66

loving

-0.65

ipop

-0.65

cffff

-0.62

chance

-0.62

onz

-0.62

POSITIVE LOGITS

 contrast

1.21

 contrasts

1.19

ly

1.05

 departure

0.93

 contradiction

0.91

 difference

0.88

 inequalities

0.88

 similarities

0.86

 stark

0.83

 disparities

0.83

Activations Density 0.103%

phrases that indicate a strong or noticeable difference

phrases emphasizing sharp contrasts or significant differences

No Comments

No Known Activations