INDEX

Explanations

geometric shapes, specifically circles

oai_token-act-pair · gpt-3.5-turbo

references to circles 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 @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ufact

-0.77

enf

-0.75

ãĥ¤

-0.73

ono

-0.72

inately

-0.71

ensional

-0.71

iary

-0.71

 Bundes

-0.69

uctions

-0.67

inen

-0.66

POSITIVE LOGITS

jer

0.96

 circle

0.88

creen

0.84

circle

0.84

hift

0.78

 jerk

0.78

house

0.78

 circles

0.78

 circling

0.78

naire

0.76

Activations Density 0.026%

geometric shapes, specifically circles

references to circles in various contexts

No Comments

No Known Activations