INDEX

Explanations

mentions of specific positions or job titles

oai_token-act-pair · gpt-3.5-turbo

references to positions or roles held by individuals or entities

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

 Qiao

-0.92

 Flavoring

-0.79

 Orchestra

-0.79

ï¸

-0.78

 Beans

-0.78

inho

-0.73

 Starr

-0.72

ãĤ¤ãĥĪ

-0.72

 Fitz

-0.71

 Thro

-0.70

POSITIVE LOGITS

gres

1.25

doctoral

1.18

mortem

1.13

hum

1.08

modern

1.01

natal

1.00

graduate

0.98

ulate

0.97

ulates

0.96

traumatic

0.93

Activations Density 0.031%

mentions of specific positions or job titles

references to positions or roles held by individuals or entities

No Comments

No Known Activations