INDEX

Explanations

words related to geographical directions, particularly specific locations such as "north", "south", "east", and "west"

oai_token-act-pair · gpt-3.5-turbo

geographical references to directions, particularly north and south

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Calculator

-0.75

 Parenthood

-0.74

ulous

-0.73

 Hacker

-0.73

amaru

-0.73

 Machina

-0.71

 Therapy

-0.71

 Unch

-0.71

 Care

-0.68

 Files

-0.68

POSITIVE LOGITS

ward

1.45

bound

1.16

wards

1.10

side

1.09

west

0.99

 coast

0.99

 pole

0.95

ampton

0.95

 west

0.95

 flank

0.94

Activations Density 0.038%

words related to geographical directions, particularly specific locations such as "north", "south", "east", and "west"

geographical references to directions, particularly north and south

No Comments

No Known Activations