INDEX

Explanations

numerical indicators of sequence or order

oai_token-act-pair · gpt-3.5-turbo

references to ordinal numbers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ¦ãĤ¹

-0.85

models

-0.72

ouls

-0.72

addons

-0.70

Engineers

-0.69

 ILCS

-0.69

Materials

-0.67

mes

-0.65

comfort

-0.65

iths

-0.65

POSITIVE LOGITS

one

1.21

two

1.13

 pair

1.03

 batch

0.98

 three

0.98

 iteration

0.91

 couple

0.91

 installment

0.88

 preceded

0.88

was

0.88

Activations Density 0.151%

numerical indicators of sequence or order

references to ordinal numbers

No Comments

No Known Activations

numerical indicators of sequence or order

references to ordinal numbers

No Comments

No Known Activations