INDEX

Explanations

descriptions of individuals such as their names and occupations

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

imation

-0.68

cean

-0.66

brids

-0.66

psc

-0.64

ibraries

-0.63

 Dragonbound

-0.63

tsy

-0.62

cients

-0.60

orea

-0.59

brid

-0.59

POSITIVE LOGITS

who

1.22

 whom

1.19

who

1.18

 whose

1.13

whose

1.12

 surn

1.07

 named

0.95

 identified

0.94

 nicknamed

0.93

 dressed

0.90

Activations Density 0.128%

descriptions of individuals such as their names and occupations

No Comments

No Known Activations