INDEX

Explanations

the word "que"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "que" and its variations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ERAL

-0.80

sonian

-0.77

auri

-0.70

ãĥŁ

-0.69

ohyd

-0.67

rition

-0.66

 scissors

-0.66

wav

-0.65

ropolitan

-0.63

ãĤ¼ãĤ¦ãĤ¹

-0.62

POSITIVE LOGITS

que

1.43

Que

1.09

erness

0.97

 queue

0.93

iu

0.86

 queues

0.85

que

0.85

 compr

0.82

 challeng

0.82

Que

0.81

Activations Density 0.006%

the word "que"

instances of the word "que" and its variations

No Comments

No Known Activations