INDEX

Explanations

numbers and figures related to losses or deficits

oai_token-act-pair · gpt-3.5-turbo

numeric data related to losses or casualties, particularly in a historical or statistical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ç«

-0.79

 GOODMAN

-0.75

sty

-0.71

 Ready

-0.70

ready

-0.69

Bey

-0.68

Ivy

-0.68

ocrine

-0.68

 bour

-0.68

Ready

-0.68

POSITIVE LOGITS

 lost

2.14

 losses

2.10

 loss

2.09

lost

1.95

 lose

1.95

 loses

1.92

loss

1.87

 losing

1.87

 Loss

1.85

Lost

1.72

Activations Density 0.246%

numbers and figures related to losses or deficits

numeric data related to losses or casualties, particularly in a historical or statistical context

No Comments

No Known Activations