INDEX

Explanations

phrases indicating a comparison of something to a specific object or concept

oai_token-act-pair · gpt-3.5-turbo

the letter "A" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

idge

-0.64

orate

-0.63

rences

-0.60

 dilig

-0.58

Interstitial

-0.58

ItemImage

-0.58

ictions

-0.56

EngineDebug

-0.54

displayText

-0.53

scape

-0.53

POSITIVE LOGITS

2.56

An

1.65

1.49

At

1.05

As

1.05

 Another

1.05

AN

1.00

One

1.00

The

0.93

Activations Density 0.050%

phrases indicating a comparison of something to a specific object or concept

the letter "A" in various contexts

No Comments

No Known Activations