INDEX

Explanations

words related to exerting influence or control over certain processes or situations

oai_token-act-pair · gpt-3.5-turbo

terms related to risk and regulation in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

avorite

-0.86

RESULTS

-0.75

 Origin

-0.66

BACK

-0.64

REE

-0.64

 Mayweather

-0.64

HOME

-0.63

GROUND

-0.63

 Grail

-0.62

WD

-0.62

POSITIVE LOGITS

izing

1.98

ization

1.94

ized

1.89

ised

1.85

izers

1.83

izations

1.81

ising

1.81

isation

1.72

izes

1.70

izable

1.67

Activations Density 0.051%

words related to exerting influence or control over certain processes or situations

terms related to risk and regulation in various contexts

No Comments

No Known Activations