INDEX

Explanations

words related to people's names

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Troitzky."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

UAL

-0.74

icago

-0.73

à¼

-0.71

,,,,

-0.70

 Reloaded

-0.69

ividual

-0.65

IST

-0.65

ãĤ¦ãĤ¹

-0.65

soever

-0.65

++++++++++++++++

-0.65

POSITIVE LOGITS

opers

1.10

dden

1.07

kefeller

1.03

tro

1.00

opa

0.98

Tro

0.96

phies

0.90

ppo

0.86

phy

0.85

pha

0.84

Activations Density 0.007%

words related to people's names

mentions of the word "Troitzky."

No Comments

No Known Activations