INDEX

Explanations

code-like strings, potentially including special characters and variable names

oai_token-act-pair · gpt-3.5-turbo

special characters or symbols often used in text encoding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

terday

-0.92

aimon

-0.73

 decomp

-0.71

 Bengal

-0.68

ifying

-0.68

ifications

-0.67

ignment

-0.66

TTL

-0.63

izable

-0.63

 Medline

-0.62

POSITIVE LOGITS

ĸ

1.45

Ķ

1.26

¹

1.25

ĺ

1.24

Ħ

1.22

ı

1.22

Ĩ

1.20

½

1.19

·

1.19

¬

1.17

Activations Density 0.012%

code-like strings, potentially including special characters and variable names

special characters or symbols often used in text encoding

No Comments

No Known Activations