INDEX

Explanations

phrases related to law enforcement actions or security measures

oai_token-act-pair · gpt-3.5-turbo

actions related to security and containment measures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

yout

-0.71

Calif

-0.67

Definition

-0.66

piracy

-0.65

-0.63

vp

-0.63

Recomm

-0.62

Policy

-0.62

TPP

-0.62

ãĤ¦ãĤ¹

-0.62

POSITIVE LOGITS

 entrances

1.44

 buildings

1.17

 residences

1.11

 rooms

1.10

 balcon

1.07

 adjoining

1.05

 roofs

1.05

 classrooms

1.04

 portions

1.03

 apartments

1.02

Activations Density 0.278%

phrases related to law enforcement actions or security measures

actions related to security and containment measures

No Comments

No Known Activations