INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names of individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

agra

-0.86

arl

-0.82

jah

-0.82

esh

-0.80

reb

-0.79

roach

-0.77

mem

-0.75

aji

-0.74

chal

-0.73

quer

-0.72

POSITIVE LOGITS

 Fowler

0.86

 Paddock

0.85

acebook

0.75

Osw

0.75

 Huntington

0.74

 Hudson

0.71

 subp

0.68

herty

0.68

 Hodg

0.67

dinand

0.66

Activations Density 0.021%

proper nouns

proper nouns, specifically names of individuals

No Comments

No Known Activations