INDEX

Explanations

dates in various formats, with a slightly higher focus on dates in May

oai_token-act-pair · gpt-3.5-turbo

date-related information

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

advertisement

-0.81

ãĥ³ãĤ¸

-0.78

atform

-0.73

ãĥ¼ãĥĨ

-0.72

rencies

-0.72

urities

-0.69

latable

-0.68

cients

-0.67

arbon

-0.66

è¦ļéĨĴ

-0.66

POSITIVE LOGITS

ouk

0.76

KH

0.68

 Cyrus

0.64

 herself

0.62

MLA

0.61

 climax

0.61

 unwitting

0.60

 seeds

0.59

Khe

0.58

hya

0.58

Activations Density 0.167%

dates in various formats, with a slightly higher focus on dates in May

date-related information

No Comments

No Known Activations