INDEX

Explanations

words related to entities or companies

oai_token-act-pair · gpt-3.5-turbo

terms related to various forms of "ity," indicating states, qualities, or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bern

-0.79

escription

-0.78

kef

-0.74

kers

-0.73

hester

-0.71

ker

-0.70

GER

-0.70

kson

-0.69

ergy

-0.69

acan

-0.68

POSITIVE LOGITS

Peb

1.02

 Principle

0.77

 quot

0.71

Pwr

0.70

 Charm

0.66

 assurance

0.66

quo

0.66

uary

0.66

 charms

0.64

ously

0.62

Activations Density 0.074%

words related to entities or companies

terms related to various forms of "ity," indicating states, qualities, or conditions

No Comments

No Known Activations