INDEX

Explanations

instances where someone is attempting to escape or avoid consequences

oai_token-act-pair · gpt-3.5-turbo

phrases about evading consequences or accountability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

wash

-0.79

tenance

-0.75

wake

-0.74

mart

-0.73

link

-0.70

she

-0.68

lehem

-0.68

scape

-0.64

stem

-0.63

main

-0.62

POSITIVE LOGITS

 impunity

0.92

 murder

0.68

 dignity

0.67

aceous

0.64

 Mankind

0.63

 unequ

0.63

 reckless

0.63

standing

0.62

 murdering

0.61

 Murder

0.60

Activations Density 0.059%

instances where someone is attempting to escape or avoid consequences

phrases about evading consequences or accountability

No Comments

No Known Activations