INDEX

Explanations

words related to typography

oai_token-act-pair · gpt-3.5-turbo

references to typography and typing terminology

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

IVERS

-0.77

IRO

-0.75

Rae

-0.69

Merit

-0.68

ITED

-0.67

 Weinstein

-0.66

å°Ĩ

-0.66

 Nero

-0.65

 Bayer

-0.64

 Levy

-0.64

POSITIVE LOGITS

typ

1.10

ology

1.08

faces

1.03

ographical

1.03

opher

0.99

otyp

0.99

ographic

0.96

riter

0.93

ographically

0.93

rist

0.93

Activations Density 0.020%

words related to typography

references to typography and typing terminology

No Known Activations