INDEX

Explanations

phrases related to circular objects or concepts

oai_token-act-pair · gpt-3.5-turbo

variations of the word "circumcision."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Ally

-0.80

Ã¡n

-0.78

resa

-0.76

Monster

-0.76

hower

-0.74

 Tillerson

-0.74

retty

-0.73

 Rover

-0.72

Ã³n

-0.70

 Patton

-0.69

POSITIVE LOGITS

circ

3.70

 Circ

3.22

 circ

3.04

Circ

2.92

irc

2.04

 circum

1.90

Cir

1.76

IRC

1.71

 circular

1.52

cir

1.48

Activations Density 0.030%

phrases related to circular objects or concepts

variations of the word "circumcision."

No Comments

No Known Activations