INDEX

Explanations

instances of doubling or significantly increasing something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating an increase or doubling of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

itives

-0.69

oria

-0.67

eller

-0.66

ogi

-0.64

RAW

-0.62

onies

-0.59

uctions

-0.59

ixtures

-0.59

rians

-0.58

istries

-0.58

POSITIVE LOGITS

 digits

1.03

 digit

0.88

plet

0.86

peak

0.73

pend

0.67

header

0.67

dig

0.67

 dipping

0.66

ouble

0.66

cedes

0.65

Activations Density 0.038%

instances of doubling or significantly increasing something

phrases indicating an increase or doubling of something

No Comments

No Known Activations