INDEX

Explanations

terms related to burglaries and theft

oai_token-act-pair · gpt-3.5-turbo

terms related to burglary and its implications

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

shr

-0.88

aft

-0.77

andr

-0.74

toggle

-0.73

nant

-0.72

alg

-0.72

inst

-0.71

LESS

-0.71

apa

-0.71

onal

-0.71

POSITIVE LOGITS

 burgl

1.30

 burglary

1.28

 burg

1.22

©¶æ¥µ

0.85

 robbery

0.81

 Burg

0.79

 alarms

0.77

 spree

0.77

 robberies

0.77

 homicide

0.75

Activations Density 0.008%

terms related to burglaries and theft

terms related to burglary and its implications

No Comments

No Known Activations