INDEX

Explanations

phrases related to prisoners or wartime situations

oai_token-act-pair · gpt-3.5-turbo

references to prisoners in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

orp

-0.84

amera

-0.76

orie

-0.71

 Boll

-0.68

alore

-0.67

wig

-0.66

Ã¼r

-0.65

laus

-0.64

uyomi

-0.63

ories

-0.63

POSITIVE LOGITS

 prisoners

1.08

 captives

0.98

 prisoner

0.92

 inmates

0.87

 detainees

0.83

 sentenced

0.80

 captive

0.79

 hostage

0.79

 hostages

0.78

 incarcerated

0.77

Activations Density 0.021%

phrases related to prisoners or wartime situations

references to prisoners in various contexts

No Comments

No Known Activations