INDEX

Explanations

named locations

oai_token-act-pair · gpt-3.5-turbo

sentences or phrases that indicate locations and directions

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 impuls

-0.91

 oath

-0.83

 aggress

-0.81

 himself

-0.81

 inconsistency

-0.80

 undecided

-0.80

 trusting

-0.79

 naive

-0.77

 reject

-0.77

 quitting

-0.76

POSITIVE LOGITS

Located

1.39

Residents

1.25

Owner

1.18

Phot

1.17

Built

1.16

Lake

1.15

Courtesy

1.12

Photo

1.11

Construction

1.09

Photos

1.07

Activations Density 0.727%

named locations

sentences or phrases that indicate locations and directions

No Known Activations