INDEX

Explanations

concepts related to breaking or entering

oai_token-act-pair · gpt-3.5-turbo

words related to breaking or entering

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

risome

-0.63

MG

-0.63

oka

-0.62

ulu

-0.61

vg

-0.58

cki

-0.58

eers

-0.58

rising

-0.57

ional

-0.56

RON

-0.56

POSITIVE LOGITS

 curfew

1.05

neck

0.92

 necks

0.89

 ribs

0.88

 into

0.85

 bones

0.83

fast

0.81

 apart

0.78

 windows

0.78

 laws

0.78

Activations Density 0.037%

concepts related to breaking or entering

words related to breaking or entering

No Comments

No Known Activations