INDEX

Explanations

names of geographical locations, particularly focused on the state of Idaho

oai_token-act-pair · gpt-3.5-turbo

mentions of the state of Idaho

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

Ü

-0.81

hops

-0.75

byter

-0.75

xual

-0.75

SHIP

-0.74

trigger

-0.72

vous

-0.71

 istg

-0.69

amins

-0.69

thood

-0.69

POSITIVE LOGITS

 Idaho

1.27

 Falls

1.20

 Potato

1.03

 Boise

1.02

 Rapids

0.89

 Nuggets

0.88

 Anchorage

0.87

 Montana

0.86

adelphia

0.84

keye

0.84

Activations Density 0.008%

names of geographical locations, particularly focused on the state of Idaho

mentions of the state of Idaho

No Comments

No Known Activations