INDEX

Explanations

mentions of interviews

oai_token-act-pair · gpt-3.5-turbo

references to interviews and laughter

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

 casc

-0.67

 charge

-0.67

amar

-0.67

 flock

-0.65

 downstream

-0.63

 upgr

-0.63

 cascade

-0.63

 blaze

-0.63

 incur

-0.62

 bearer

-0.61

POSITIVE LOGITS

Interview

3.20

 Interview

3.03

laughs

1.93

Laughs

1.66

laugh

1.32

 interviewing

1.20

 interview

1.10

 interviews

1.09

 rehears

1.06

laughter

1.06

Activations Density 0.048%

mentions of interviews

references to interviews and laughter

No Comments

No Known Activations