INDEX

Explanations

phrases related to comparing or positioning individuals or groups in terms of status or influence

oai_token-act-pair · gpt-3.5-turbo

references to groups or communities

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

ossession

-0.85

odder

-0.72

idation

-0.70

oire

-0.69

ober

-0.68

aura

-0.67

oldemort

-0.64

dayName

-0.63

emb

-0.63

oul

-0.62

POSITIVE LOGITS

 peers

0.97

 those

0.94

 them

0.91

 equals

0.88

 classmates

0.88

 contemporaries

0.85

 strangers

0.84

 colleagues

0.83

 ourselves

0.82

friends

0.82

Activations Density 0.054%

phrases related to comparing or positioning individuals or groups in terms of status or influence

references to groups or communities

No Comments

No Known Activations