INDEX

Explanations

locations based on the context of the document

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etheless

-0.70

ylum

-0.67

 diarr

-0.64

 charact

-0.63

igan

-0.63

 monop

-0.63

 parasite

-0.62

VALUE

-0.62

yip

-0.62

 contrace

-0.61

POSITIVE LOGITS

 County

0.84

 Residents

0.82

 Wells

0.76

 Unified

0.75

 Clippers

0.75

 Rapids

0.75

 COUNTY

0.75

 Liberties

0.73

area

0.73

edu

0.71

Activations Density 0.054%

locations based on the context of the document

No Comments

No Known Activations