INDEX

Explanations

situations where a preference is expressed

oai_token-act-pair · gpt-3.5-turbo

expressions of preferences or choices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

pack

-0.77

Haunted

-0.76

idem

-0.69

bleacher

-0.68

MD

-0.66

ammy

-0.65

Tour

-0.64

akening

-0.64

Article

-0.63

ese

-0.63

POSITIVE LOGITS

 preferring

0.98

yip

0.91

 prefers

0.85

 prefer

0.85

lihood

0.80

 warr

0.80

quickShipAvailable

0.77

swer

0.77

 preferred

0.77

 favoured

0.75

Activations Density 0.008%

situations where a preference is expressed

expressions of preferences or choices

No Comments

No Known Activations