INDEX

Explanations

instances of the word "or" with high activation values

oai_token-act-pair · gpt-3.5-turbo

the word "or" and its usage in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

mostly

-0.79

then

-0.65

now

-0.65

tackle

-0.64

probably

-0.59

NOW

-0.58

Lots

-0.58

Probably

-0.57

Almost

-0.56

Here

-0.56

POSITIVE LOGITS

 anything

1.54

any

1.24

 anywhere

1.18

 anybody

1.13

 slightest

1.12

 anyone

1.11

 even

1.10

chard

1.10

 anymore

1.09

 whatsoever

1.09

Activations Density 0.093%

instances of the word "or" with high activation values

the word "or" and its usage in various contexts

No Comments

No Known Activations