INDEX

Explanations

objects or actions related to chairs

oai_token-act-pair · gpt-3.5-turbo

references to chairs and seating arrangements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Territories

-0.76

aries

-0.67

spect

-0.65

 Partnership

-0.63

film

-0.62

 Rounds

-0.62

actionDate

-0.62

 Adin

-0.61

via

-0.61

uder

-0.61

POSITIVE LOGITS

lift

1.36

 chair

1.16

 chairs

1.09

manship

1.07

chair

1.02

woman

1.02

holder

1.01

tops

0.99

Desk

0.98

chairs

0.95

Activations Density 0.015%

objects or actions related to chairs

references to chairs and seating arrangements

No Comments

No Known Activations