INDEX

Explanations

numerical patterns or sequences in the document

oai_token-act-pair · gpt-3.5-turbo

numerical identifiers or references in a structured format

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

nesday

-0.81

pire

-0.78

 Flavoring

-0.75

itionally

-0.72

achus

-0.72

 proble

-0.70

spread

-0.68

parency

-0.68

 disadvant

-0.68

iya

-0.66

POSITIVE LOGITS

dfx

0.94

0.87

0.85

ogram

0.80

mph

0.79

obal

0.78

ographically

0.78

obyl

0.76

 Minutes

0.76

othing

0.73

Activations Density 0.051%

numerical patterns or sequences in the document

numerical identifiers or references in a structured format

No Comments

No Known Activations