INDEX

Explanations

phrases related to detecting various elements or conditions

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of detection

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

 McKin

-0.72

 Conference

-0.68

OUR

-0.66

 mortg

-0.62

 conference

-0.62

 Schro

-0.61

 Finance

-0.60

USA

-0.60

 Salary

-0.59

 Story

-0.59

POSITIVE LOGITS

 detect

3.55

 detecting

2.47

 detects

2.32

 detected

2.18

 detection

2.16

Detect

1.90

 Detect

1.89

 Detection

1.86

 detectable

1.56

 detector

1.49

Activations Density 0.008%

phrases related to detecting various elements or conditions

terms related to the concept of detection

No Comments

No Known Activations