INDEX

Explanations

words related to specific locations, particularly street names or addresses

oai_token-act-pair · gpt-3.5-turbo

words related to respiratory and operational processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

iban

-0.93

plates

-0.79

ŃĶ

-0.75

iple

-0.69

roman

-0.69

phabet

-0.69

ako

-0.68

teen

-0.66

auri

-0.66

forums

-0.65

POSITIVE LOGITS

pload

0.76

resent

0.76

Ö¼

0.75

 Rodham

0.73

ionage

0.71

vous

0.69

resp

0.68

ocaly

0.68

×¨

0.67

 resp

0.67

Activations Density 0.025%

words related to specific locations, particularly street names or addresses

words related to respiratory and operational processes

No Comments

No Known Activations