INDEX

Explanations

phrases or words related to regulatory actions or control

oai_token-act-pair · gpt-3.5-turbo

terminology related to regulation and governance

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Plane

-0.70

 Falcons

-0.70

lihood

-0.68

 Saints

-0.68

 Jian

-0.68

 Apostle

-0.67

spell

-0.67

 Ashes

-0.67

Inv

-0.67

stone

-0.66

POSITIVE LOGITS

 regulate

1.23

 regulates

1.19

 regulated

1.12

 regulating

1.08

ategory

1.00

 unregulated

0.96

 regulator

0.96

untary

0.95

regulated

0.94

untarily

0.94

Activations Density 0.008%

phrases or words related to regulatory actions or control

terminology related to regulation and governance

No Comments

No Known Activations