INDEX

Explanations

references to individuals detained in specific contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to detainees and their conditions

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

bc

-0.80

jri

-0.72

nect

-0.71

bye

-0.70

bye

-0.68

orsche

-0.67

abeth

-0.66

ardon

-0.66

uncture

-0.66

#$

-0.64

POSITIVE LOGITS

 detainees

1.06

anamo

0.96

 captives

0.78

 facilities

0.78

rans

0.77

']

0.74

 detainee

0.74

 detained

0.73

 seekers

0.72

 langu

0.69

Activations Density 0.015%

references to individuals detained in specific contexts

terms related to detainees and their conditions

No Comments

No Known Activations