INDEX

Explanations

phrases related to trends or movements going upwards

oai_token-act-pair · gpt-3.5-turbo

terms indicating trends, particularly upward trends

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mong

-0.77

tein

-0.74

 Bigfoot

-0.72

 Mamm

-0.71

 Bloody

-0.68

TOP

-0.67

BuyableInstoreAndOnline

-0.66

BILITY

-0.65

 Lithuan

-0.64

DN

-0.64

POSITIVE LOGITS

¥ŀ

1.18

srf

1.01

 entreprene

0.97

 trajectory

0.90

 proport

0.89

grading

0.89

 traject

0.89

 earthqu

0.87

 upward

0.86

graded

0.86

Activations Density 0.006%

phrases related to trends or movements going upwards

terms indicating trends, particularly upward trends

No Comments

No Known Activations