INDEX

Explanations

phrases related to giving advice or instructions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating suggestions or recommendations

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

 argues

-0.84

 acknowledges

-0.67

 contends

-0.67

 advocates

-0.65

 asserts

-0.65

 cite

-0.64

 ensures

-0.63

ths

-0.63

 Polit

-0.62

 cites

-0.62

POSITIVE LOGITS

apest

0.77

orage

0.71

—"

0.70

âĢ¦"

0.70

 âĢ¦"

0.69

âĢ¦."

0.66

 mosqu

0.64

 conflic

0.63

 fitt

0.62

 prank

0.62

Activations Density 1.637%

phrases related to giving advice or instructions

phrases indicating suggestions or recommendations

No Comments

No Known Activations