INDEX

Explanations

words related to law enforcement agencies

oai_token-act-pair · gpt-3.5-turbo

references to law enforcement or the police

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

ppo

-0.96

hold

-0.93

bourg

-0.80

ansas

-0.74

glers

-0.73

holding

-0.71

ishi

-0.71

hab

-0.68

fram

-0.68

itud

-0.68

POSITIVE LOGITS

PD

1.00

GF

0.95

iamond

0.93

illon

0.89

ouble

0.86

insula

0.85

oodle

0.85

ective

0.80

ION

0.77

uterte

0.77

Activations Density 0.010%

words related to law enforcement agencies

references to law enforcement or the police

No Comments

No Known Activations