INDEX

Explanations

references to numbers, specifically those related to quantities or measurements

oai_token-act-pair · gpt-3.5-turbo

occurrences of the number 1000

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Downloadha

-0.87

reon

-0.87

Reviewer

-0.87

unal

-0.75

gettable

-0.74

 commanding

-0.74

gotten

-0.73

hap

-0.73

blem

-0.72

olulu

-0.68

POSITIVE LOGITS

0.95

mAh

0.90

0.88

0.87

MQ

0.85

0.83

 msec

0.82

ãĤ¨ãĥ«

0.80

ÙĨ

0.80

Activations Density 0.011%

references to numbers, specifically those related to quantities or measurements

occurrences of the number 1000

No Comments

No Known Activations