INDEX

Explanations

dates represented as numerical values followed by a th, nd, or rd suffix

oai_token-act-pair · gpt-3.5-turbo

dates and numerical information related to events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

attled

-0.66

faced

-0.63

rw

-0.62

akespe

-0.62

orable

-0.59

 Genius

-0.58

chwitz

-0.57

 cheated

-0.57

edIn

-0.56

father

-0.55

POSITIVE LOGITS

th

1.07

TH

0.80

ths

0.77

Interstitial

0.74

rd

0.73

NESS

0.71

teenth

0.69

aos

0.68

ECA

0.68

ESE

0.67

Activations Density 0.102%

dates represented as numerical values followed by a th, nd, or rd suffix

dates and numerical information related to events

No Comments

No Known Activations