INDEX

Explanations

words related to prizes or rewards

oai_token-act-pair · gpt-3.5-turbo

references to prizes and their significance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

heads

-0.81

 Hurricanes

-0.81

tracks

-0.78

 Leopard

-0.75

 Leilan

-0.74

ocker

-0.74

head

-0.74

ãĤ¨ãĥ«

-0.72

 Situation

-0.71

enegger

-0.70

POSITIVE LOGITS

pri

1.24

etary

1.05

eties

1.03

eteen

1.00

ety

0.95

eteenth

0.91

hon

0.91

İ

0.90

¦

0.87

ĸļ

0.86

Activations Density 0.005%

words related to prizes or rewards

references to prizes and their significance

No Comments

No Known Activations