INDEX

Explanations

instances of questions starting with "What's" or "What is"

oai_token-act-pair · gpt-3.5-turbo

phrases that include the character sequence "Ļ" and similar looking patterns

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

UV

-0.67

 dwindling

-0.65

 circulation

-0.64

 Mobil

-0.64

 boarding

-0.62

 duly

-0.61

toe

-0.61

 sprawling

-0.61

 Trinidad

-0.59

 flowering

-0.59

POSITIVE LOGITS

else

1.21

different

1.02

matter

1.01

these

0.99

everyone

0.95

should

0.95

mean

0.94

they

0.93

requires

0.93

you

0.92

Activations Density 0.132%

instances of questions starting with "What's" or "What is"

phrases that include the character sequence "Ļ" and similar looking patterns

No Comments

No Known Activations