INDEX

Explanations

words related to geographical locations or places

oai_token-act-pair · gpt-3.5-turbo

instances of the string "ap" or variations of it within different contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Monstrous

-0.64

 silence

-0.63

 phyl

-0.62

IGHTS

-0.60

 FANT

-0.60

 omission

-0.59

 interf

-0.58

 Dragons

-0.58

tty

-0.57

 explan

-0.57

POSITIVE LOGITS

olitan

1.42

ropri

1.28

ocalypse

1.28

oleon

1.24

ocalyptic

1.19

rison

1.14

olitics

1.09

acity

1.07

edia

1.07

rote

1.06

Activations Density 0.035%

words related to geographical locations or places

instances of the string "ap" or variations of it within different contexts

No Comments

No Known Activations