INDEX

Explanations

polygon or geometric shape descriptors

oai_token-act-pair · gpt-3.5-turbo

descriptive terms for geometric shapes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Stew

-0.73

imester

-0.71

rolog

-0.70

skill

-0.68

tested

-0.68

ufact

-0.67

elf

-0.67

 Younger

-0.67

INS

-0.66

learn

-0.66

POSITIVE LOGITS

 shaped

1.27

shaped

1.15

 triangular

1.07

 rectangular

1.06

 oval

1.04

 shape

1.02

cyl

1.01

 ellipt

0.99

 shapes

0.93

 spherical

0.93

Activations Density 0.066%

polygon or geometric shape descriptors

descriptive terms for geometric shapes

No Comments

No Known Activations