INDEX

Explanations

descriptions related to geographical locations and amenities

oai_token-act-pair · gpt-3.5-turbo

descriptive phrases about scenic locations and views

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

imester

-0.66

Consumer

-0.65

ealous

-0.61

olicited

-0.61

 Reader

-0.60

 Lifetime

-0.59

 Ronnie

-0.59

aughs

-0.59

 Parent

-0.58

obin

-0.58

POSITIVE LOGITS

 limestone

1.22

 cliffs

1.16

 coastline

1.13

 cavern

1.12

 hills

1.09

 caves

1.08

 ruins

1.03

 canyon

1.03

 buildings

1.02

 mountainous

1.02

Activations Density 1.268%

descriptions related to geographical locations and amenities

descriptive phrases about scenic locations and views

No Comments

No Known Activations