INDEX

Explanations

proper nouns or entities related to specific places or people

oai_token-act-pair · gpt-3.5-turbo

the word "amp" or variations thereof, likely referring to amplification or amperage in a technical context

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

ĪĴ

-0.88

SHIP

-0.77

everal

-0.74

 STATES

-0.66

zanne

-0.66

gian

-0.64

cki

-0.64

 obser

-0.64

DEM

-0.63

interstitial

-0.62

POSITIVE LOGITS

shire

1.12

aign

1.03

agne

1.00

hetamine

0.96

oons

0.90

oline

0.88

eters

0.87

aic

0.86

stead

0.85

agn

0.84

Activations Density 0.009%

proper nouns or entities related to specific places or people

the word "amp" or variations thereof, likely referring to amplification or amperage in a technical context

No Comments

No Known Activations