INDEX

Explanations

the word "que" repeated throughout the text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "que"

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

ERAL

-0.92

ãĥŁ

-0.79

ortium

-0.74

ãĥĩ

-0.73

sonian

-0.72

auri

-0.71

ramid

-0.71

ãĤ¼ãĤ¦ãĤ¹

-0.71

RAL

-0.69

cipline

-0.68

POSITIVE LOGITS

que

1.17

Que

1.05

iu

0.90

erness

0.86

uing

0.85

ued

0.81

fue

0.78

que

0.77

ues

0.76

 compr

0.72

Activations Density 0.004%

the word "que" repeated throughout the text

instances of the word "que"

No Comments

No Known Activations