INDEX

Explanations

phrases related to causation and consequences

oai_token-act-pair · gpt-3.5-turbo

phrases indicating causes and effects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

DragonMagazine

-0.86

enth

-0.72

soon

-0.71

jan

-0.70

Rated

-0.69

itsch

-0.69

ister

-0.68

itus

-0.67

uristic

-0.67

alian

-0.66

POSITIVE LOGITS

 fewer

1.10

 decreased

1.07

 increased

1.04

 considerable

1.00

 reduced

0.98

 drastic

0.98

 significant

0.98

 substantial

0.98

 tighter

0.97

 greater

0.96

Activations Density 0.119%

phrases related to causation and consequences

phrases indicating causes and effects

No Comments

No Known Activations