INDEX

Explanations

names of specific locations or people

oai_token-act-pair · gpt-3.5-turbo

the word "har" and its variations used in different contexts

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

psey

-0.81

 transitional

-0.78

BOOK

-0.73

htaking

-0.72

inition

-0.72

 detector

-0.70

 nomine

-0.70

 tremend

-0.70

 dotted

-0.68

 sheet

-0.66

POSITIVE LOGITS

vest

1.10

har

1.09

vard

1.04

assment

1.02

rod

1.01

thal

1.00

vati

0.99

MED

0.98

riet

0.95

Har

0.92

Activations Density 0.005%

names of specific locations or people

the word "har" and its variations used in different contexts

No Comments

No Known Activations