INDEX

Explanations

geometric shapes, particularly circles

oai_token-act-pair · gpt-3.5-turbo

references to circles or circular formations in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ãĥ¤

-0.76

ensional

-0.66

ractive

-0.65

ainer

-0.65

NF

-0.64

icient

-0.63

determination

-0.62

inated

-0.62

enf

-0.61

ressive

-0.61

POSITIVE LOGITS

 circle

1.24

circle

1.12

 circles

0.95

naire

0.92

wheel

0.90

 Dise

0.87

 circling

0.84

 Circle

0.84

cules

0.81

SHIP

0.78

Activations Density 0.005%

geometric shapes, particularly circles

references to circles or circular formations in various contexts

No Comments

No Known Activations