INDEX

Explanations

variations or differences in characteristics or attributes within a set of related items or concepts

oai_token-act-pair · gpt-3.5-turbo

phrases indicating variability or differences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

odore

-0.78

zan

-0.74

Ķ

-0.70

oops

-0.70

°

-0.66

worms

-0.66

İ

-0.64

advertising

-0.62

TE

-0.62

gone

-0.62

POSITIVE LOGITS

 wildly

1.27

 depending

1.23

 greatly

1.15

 considerably

1.13

 widely

1.11

 markedly

1.10

 enormously

1.05

 tremendously

1.05

 drastically

1.03

 geographically

1.02

Activations Density 0.063%

variations or differences in characteristics or attributes within a set of related items or concepts

phrases indicating variability or differences

No Comments

No Known Activations