INDEX

Explanations

phrases related to staying or residing in different locations

oai_token-act-pair · gpt-3.5-turbo

phrases related to temporary living situations or accommodations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ggles

-0.67

ocrat

-0.63

ĸļ

-0.63

pex

-0.62

phony

-0.59

catentry

-0.59

 fielded

-0.58

Ĳ

-0.56

 promulg

-0.56

ulus

-0.56

POSITIVE LOGITS

 aboard

1.04

 upstairs

1.00

 downstairs

0.99

 indoors

0.96

 abroad

0.93

 quarters

0.93

 inside

0.93

 ashore

0.92

 elsewhere

0.86

 nearby

0.85

Activations Density 0.287%

phrases related to staying or residing in different locations

phrases related to temporary living situations or accommodations

No Known Activations