INDEX

Explanations

words related to names and locations

oai_token-act-pair · gpt-3.5-turbo

occurrences of specific names or characters, particularly related to a notable figure or term

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DonaldTrump

-0.79

bread

-0.64

idious

-0.63

ulators

-0.62

 suppress

-0.62

utterstock

-0.62

orial

-0.62

ifier

-0.62

hander

-0.62

itious

-0.61

POSITIVE LOGITS

Ã¸

1.19

Ã¥

0.95

¶

0.88

 Andersen

0.87

hett

0.87

ĨĴ

0.84

ĺ

0.84

Ã¦

0.83

borg

0.83

Å¡

0.81

Activations Density 0.008%

words related to names and locations

occurrences of specific names or characters, particularly related to a notable figure or term

No Comments

No Known Activations