INDEX

Explanations

proper nouns and locations

oai_token-act-pair · gpt-3.5-turbo

references to significant elements or themes in a narrative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

culosis

-0.90

 Regarding

-0.87

 Including

-0.85

 Therefore

-0.84

[+

-0.80

ajor

-0.80

DT

-0.80

ICE

-0.79

termination

-0.78

TN

-0.78

POSITIVE LOGITS

 dazzling

1.33

 towering

1.31

 enigmatic

1.27

 uncanny

1.26

 sleek

1.26

 eerie

1.26

 bleak

1.22

 shadowy

1.22

 sprawling

1.22

 stark

1.20

Activations Density 0.519%

proper nouns and locations

references to significant elements or themes in a narrative context

No Comments

No Known Activations