INDEX

Explanations

terms related to legal actions and punitive measures, particularly focusing on the concept of imprisonment

oai_token-act-pair · gpt-3.5-turbo

terms related to imprisonment and incarceration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

cler

-0.77

thodox

-0.71

ergy

-0.68

soDeliveryDate

-0.66

Diff

-0.64

 Diff

-0.60

hy

-0.60

opic

-0.60

stage

-0.59

 stake

-0.59

POSITIVE LOGITS

 incarcer

1.18

 imprison

1.14

 imprisonment

1.13

arcer

1.09

 incarceration

1.06

 detain

0.97

 incarcerated

0.95

 imprisoned

0.92

 convict

0.77

ante

0.76

Activations Density 0.010%

terms related to legal actions and punitive measures, particularly focusing on the concept of imprisonment

terms related to imprisonment and incarceration

No Comments

No Known Activations