INDEX

Explanations

words related to liking or preference

oai_token-act-pair · gpt-3.5-turbo

expressions of preference or liking

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

arta

-1.04

 enthusi

-0.82

lehem

-0.81

ilion

-0.72

PATH

-0.72

vernment

-0.71

tek

-0.70

ureau

-0.70

useum

-0.70

ourse

-0.70

POSITIVE LOGITS

lihood

1.20

ably

0.98

 watching

0.80

lier

0.79

 seeing

0.79

liness

0.77

liest

0.73

 surprises

0.72

ability

0.68

 hearing

0.67

Activations Density 0.064%

words related to liking or preference

expressions of preference or liking

No Comments

No Known Activations