INDEX

Explanations

phrases denoting actions that one person can take on behalf of another person

oai_token-act-pair · gpt-3.5-turbo

phrases that express capability or potential actions by subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 understatement

-0.75

 irony

-0.74

 contraction

-0.70

 downside

-0.69

 negatives

-0.67

alan

-0.66

 caveats

-0.65

pse

-0.63

etheless

-0.62

 lately

-0.61

POSITIVE LOGITS

sylv

0.91

vre

0.82

ctic

0.72

 satisfy

0.72

 maximize

0.71

Ô

0.71

 survive

0.70

 continue

0.69

 safely

0.68

 regain

0.68

Activations Density 0.194%

phrases denoting actions that one person can take on behalf of another person

phrases that express capability or potential actions by subjects

No Comments

No Known Activations