INDEX

Explanations

terms related to negative experiences involving criticism or mistreatment

oai_token-act-pair · gpt-3.5-turbo

terms related to verbal abuse or derogatory behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

auga

-0.84

metics

-0.78

 nomine

-0.76

hemy

-0.75

ertodd

-0.71

profit

-0.70

 fortun

-0.69

ramid

-0.69

opsy

-0.68

urrection

-0.68

POSITIVE LOGITS

ber

1.22

keley

0.93

 Laden

0.72

atory

0.72

rill

0.70

liner

0.70

 aloud

0.69

erb

0.69

ating

0.68

hof

0.67

Activations Density 0.018%

terms related to negative experiences involving criticism or mistreatment

terms related to verbal abuse or derogatory behavior

No Comments

No Known Activations