INDEX

Explanations

words starting with the letter 'n' followed by a number, potentially indicating numerical values in a technical context

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'n'

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

single token corresponding to ' n'

Explanation Uploaded by User

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 frank

-0.63

 slack

-0.61

 compe

-0.58

 Remastered

-0.58

 trophy

-0.56

 WRITE

-0.56

Vaj

-0.56

 welf

-0.56

 Quartz

-0.55

 Shard

-0.55

POSITIVE LOGITS

autical

1.34

ativity

1.34

ameless

1.30

imble

1.25

uggets

1.25

ectar

1.23

anny

1.17

ginx

1.17

intendo

1.16

ukes

1.16

Activations Density 0.025%

words starting with the letter 'n' followed by a number, potentially indicating numerical values in a technical context

occurrences of the letter 'n'

single token corresponding to ' n'

No Comments

No Known Activations