INDEX

Explanations

preliminary information or plans

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "preliminary."

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

 Forever

-0.69

 forever

-0.66

 EVER

-0.66

 Sharing

-0.65

Ds

-0.65

 thrive

-0.64

 effortlessly

-0.64

osphere

-0.62

 Never

-0.62

 Everywhere

-0.61

POSITIVE LOGITS

 preliminary

3.60

 prelim

2.12

 Prel

2.12

iminary

2.09

 provisional

1.82

 tentative

1.66

 prepar

1.39

 initial

1.30

initial

1.15

Initial

1.14

Activations Density 0.017%

preliminary information or plans

occurrences of the word "preliminary."

No Comments

No Known Activations