INDEX

Explanations

terms related to geographical locations and entities

oai_token-act-pair · gpt-3.5-turbo

symbols or markings that indicate emotional or accentuated expressions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 "...

-0.98

"...

-0.98

...

-0.87

...?

-0.85

,...

-0.72

--

-0.71

...)

-0.69

 pokemon

-0.69

Guest

-0.69

 alot

-0.68

POSITIVE LOGITS

 ÂŃ

4.37

ÂŃ

2.27

âĢĲ

1.78

âĢĳ

1.76

————————

1.63

 [âĢ¦]

1.56

ÃĤ

1.52

 âĢ¦"

1.40

1.37

âĢ¢

1.27

Activations Density 0.015%

terms related to geographical locations and entities

symbols or markings that indicate emotional or accentuated expressions

No Comments

No Known Activations