INDEX

Explanations

expressions or mentions of regret

oai_token-act-pair · gpt-3.5-turbo

expressions of regret

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ammy

-0.78

ipt

-0.71

ymph

-0.66

 dotted

-0.66

arnaev

-0.66

IPS

-0.65

chool

-0.63

iop

-0.62

 Appalachian

-0.61

IP

-0.61

POSITIVE LOGITS

 regrets

1.30

 regret

1.25

 regretted

1.12

fully

1.00

 remorse

0.90

imaru

0.84

 gratification

0.84

ful

0.84

fulness

0.82

faced

0.81

Activations Density 0.006%

expressions or mentions of regret

expressions of regret

No Comments

No Known Activations