INDEX

Explanations

terms related to self-destructive behavior and actions

oai_token-act-pair · gpt-3.5-turbo

references to self-destructive behavior and its consequences

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

icago

-0.73

ãĥīãĥ©ãĤ´ãĥ³

-0.69

mx

-0.68

ushi

-0.66

uits

-0.65

inho

-0.64

Ã¡n

-0.63

 squat

-0.63

_-

-0.62

ayan

-0.62

POSITIVE LOGITS

destruct

2.96

Lys

1.89

 destruct

1.43

 Columb

1.36

sufficient

0.99

 Foss

0.96

 Judith

0.91

CTV

0.83

heartedly

0.82

existent

0.80

Activations Density 0.038%

terms related to self-destructive behavior and actions

references to self-destructive behavior and its consequences

No Comments

No Known Activations