INDEX

Explanations

questions

oai_token-act-pair · gpt-3.5-turbo

questions that raise inquiries or seek clarification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arm

-0.74

ikuman

-0.68

yak

-0.66

elist

-0.66

 corrid

-0.65

eared

-0.64

dex

-0.64

 sausage

-0.62

 coil

-0.60

armed

-0.60

POSITIVE LOGITS

 Surely

1.20

 Wouldn

1.20

 Answer

1.17

Answer

1.12

 Certainly

1.06

 Would

1.03

Isn

1.02

 Nope

1.01

 Perhaps

1.01

 Well

1.01

Activations Density 0.112%

questions

questions that raise inquiries or seek clarification

No Comments

No Known Activations