INDEX

Explanations

names of geographical locations and universities

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

GER

-0.77

DM

-0.76

ECH

-0.74

bm

-0.72

IJ

-0.72

icon

-0.71

BM

-0.71

gence

-0.71

CG

-0.70

DI

-0.70

POSITIVE LOGITS

ãĤº

0.77

 Santorum

0.75

 Rutgers

0.73

 Rolling

0.71

Roe

0.70

 Presbyter

0.69

 Salam

0.69

 Samson

0.68

 Starfleet

0.68

 Pulitzer

0.68

Activations Density 0.185%

names of geographical locations and universities

No Comments

No Known Activations