INDEX

Explanations

phrases related to a specific location, specifically 'Champaign'

oai_token-act-pair · gpt-3.5-turbo

references to a specific place or location

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

ias

-0.76

romeda

-0.75

Har

-0.74

forge

-0.72

Material

-0.71

zos

-0.71

stan

-0.70

 Uzbek

-0.70

arry

-0.68

 Orion

-0.67

POSITIVE LOGITS

 Champ

1.97

Champ

1.95

 cone

1.66

 Chick

1.62

 cones

1.53

CL

1.42

 champ

1.36

 Mick

1.26

 Chun

1.23

 Joan

1.19

Activations Density 0.053%

phrases related to a specific location, specifically 'Champaign'

references to a specific place or location

No Comments

No Known Activations