INDEX

Explanations

locations, particularly with a directional aspect differentiating between north and south

oai_token-act-pair · gpt-3.5-turbo

geographical references related to northern and southern locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Documents

-0.75

 Calculator

-0.74

 Files

-0.71

 Meal

-0.71

 Injury

-0.71

Deal

-0.69

 Unch

-0.69

 Device

-0.68

 Hacker

-0.68

ulous

-0.68

POSITIVE LOGITS

ward

1.40

bound

1.15

side

1.10

wards

1.07

 hemisphere

0.98

 pole

0.95

ampton

0.91

 east

0.90

 west

0.90

west

0.90

Activations Density 0.035%

locations, particularly with a directional aspect differentiating between north and south

geographical references related to northern and southern locations

No Comments

No Known Activations