INDEX

Explanations

words related to numerical values and measurements

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "digit" followed by numerical values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

erker

-0.78

 Shield

-0.75

 Cross

-0.66

 Rapp

-0.66

 Inner

-0.65

 Whole

-0.64

 Mara

-0.62

 Rath

-0.61

sterdam

-0.61

hire

-0.60

POSITIVE LOGITS

digit

1.00

 ILCS

0.93

taining

0.91

oded

0.90

omial

0.89

itized

0.89

istors

0.86

itive

0.86

ipher

0.83

rolet

0.82

Activations Density 0.005%

words related to numerical values and measurements

occurrences of the term "digit" followed by numerical values

No Comments

No Known Activations