INDEX

Explanations

phrases related to specific geographic locations, specifically focusing on locations in New England and Mexico

oai_token-act-pair · gpt-3.5-turbo

mentions of the regions "New England" and "New Mexico."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

VALUE

-0.74

ãĤ¼ãĤ¦ãĤ¹

-0.72

Bir

-0.63

BALL

-0.63

Hug

-0.63

 Mehran

-0.62

 abroad

-0.60

beg

-0.60

ãĤ¨ãĥ«

-0.60

metadata

-0.59

POSITIVE LOGITS

essee

0.93

eers

0.88

phrine

0.76

idia

0.71

aily

0.71

lear

0.70

leans

0.70

uctor

0.69

anwhile

0.69

scill

0.68

Activations Density 0.039%

phrases related to specific geographic locations, specifically focusing on locations in New England and Mexico

mentions of the regions "New England" and "New Mexico."

No Comments

No Known Activations