INDEX

Explanations

dates or years within the text

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly commas

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

pac

-0.74

Higher

-0.72

MpServer

-0.70

UGC

-0.69

Detailed

-0.69

Optional

-0.69

Matrix

-0.67

Ķ

-0.65

int

-0.63

pec

-0.63

POSITIVE LOGITS

it

0.95

 nobody

0.90

 there

0.90

 they

0.84

 pundits

0.83

he

0.82

 everyone

0.82

 many

0.81

 tensions

0.79

 rumors

0.78

Activations Density 0.176%

dates or years within the text

punctuation marks, particularly commas

No Comments

No Known Activations