INDEX

Explanations

statements about conducting interviews

oai_token-act-pair · gpt-3.5-turbo

instances of the word "interview."

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

forces

-0.69

 Nadu

-0.69

WARE

-0.68

peria

-0.65

ensions

-0.64

ides

-0.63

Veh

-0.63

atures

-0.62

 trillions

-0.61

wre

-0.61

POSITIVE LOGITS

 interview

3.64

 interviews

2.79

 Interview

2.70

 interviewer

2.58

Interview

2.40

 interviewing

2.20

 interviewed

2.07

 interrogation

1.38

 questionnaire

1.38

 audition

1.34

Activations Density 0.017%

statements about conducting interviews

instances of the word "interview."

No Comments

No Known Activations