INDEX

Explanations

phrases indicating a strong positive outcome or response

oai_token-act-pair · gpt-3.5-turbo

instances of the word "resound" and its variations, indicating strong emphasis or agreement

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Romans

-0.78

 Welsh

-0.71

 Galile

-0.64

 Catalan

-0.63

 Jobs

-0.62

 spoilers

-0.61

 Holmes

-0.60

 Turks

-0.60

 Masters

-0.59

BALL

-0.59

POSITIVE LOGITS

ounding

1.32

ourced

1.25

olute

1.23

umption

1.21

olutely

1.19

pec

1.18

uming

1.12

olver

1.12

ourcing

1.11

itance

1.10

Activations Density 0.011%

phrases indicating a strong positive outcome or response

instances of the word "resound" and its variations, indicating strong emphasis or agreement

No Comments

No Known Activations