INDEX

Explanations

phrases indicating where content can be accessed or reached

oai_token-act-pair · gpt-3.5-turbo

phrases related to audio or locating information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 vaccinations

-0.71

uked

-0.71

 clipboard

-0.70

 typh

-0.69

 evacuation

-0.68

 burial

-0.67

 antidepressants

-0.65

 disposal

-0.64

 diagnoses

-0.62

 anesthesia

-0.62

POSITIVE LOGITS

ById

0.81

 HERE

0.78

 amidst

0.72

(),

0.72

nikov

0.72

lopp

0.71

 separately

0.70

uti

0.69

erala

0.69

resso

0.68

Activations Density 0.062%

phrases indicating where content can be accessed or reached

phrases related to audio or locating information

No Comments

No Known Activations