INDEX

Explanations

numbers with possible time expressions like years, months, and hours

oai_token-act-pair · gpt-3.5-turbo

numerical values indicating timeframes or quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

actionDate

-0.77

 Caption

-0.74

 Hilbert

-0.69

£ı

-0.67

 Manila

-0.60

 underside

-0.59

ifice

-0.58

 guid

-0.57

 Strait

-0.57

reference

-0.57

POSITIVE LOGITS

 trillion

0.94

aciously

0.92

yrs

0.91

 bucks

0.90

 years

0.90

years

0.89

+)

0.89

 thousand

0.85

Chan

0.85

+,

0.84

Activations Density 0.332%

numbers with possible time expressions like years, months, and hours

numerical values indicating timeframes or quantities

No Comments

No Known Activations