INDEX

Explanations

medical terms or advanced vocabulary

oai_token-act-pair · gpt-3.5-turbo

terms related to advancements and negative actions or conditions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SIZE

-0.73

 Kinnikuman

-0.70

gob

-0.70

LOAD

-0.70

liness

-0.67

mie

-0.67

hare

-0.65

âĶģ

-0.65

 Truman

-0.65

dar

-0.64

POSITIVE LOGITS

anced

1.31

ocate

1.16

ancing

1.13

ances

1.03

adv

1.00

Adv

0.96

ices

0.93

Adv

0.92

ocated

0.91

igmat

0.90

Activations Density 0.012%

medical terms or advanced vocabulary

terms related to advancements and negative actions or conditions

No Comments

No Known Activations