INDEX

Explanations

words related to specific languages and characters

oai_token-act-pair · gpt-3.5-turbo

specific non-English characters or symbols

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

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

imentary

-0.71

swick

-0.68

nesday

-0.68

ifference

-0.66

SON

-0.66

 Jelly

-0.64

minster

-0.64

olphin

-0.63

 Donation

-0.62

aday

-0.62

POSITIVE LOGITS

Į

1.84

©

1.70

¹

1.64

Ľ

1.61

Ķ

1.61

½

1.61

ħ

1.60

¾

1.60

Ļ

1.57

¼

1.57

Activations Density 0.017%

words related to specific languages and characters

specific non-English characters or symbols

No Comments

No Known Activations