INDEX

Explanations

mentions of the ability to do specific actions or tasks

oai_token-act-pair · gpt-3.5-turbo

repetitive use of the word "can."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Lauder

-0.67

quartered

-0.64

 Federation

-0.63

 Acquisition

-0.60

ele

-0.60

 vacated

-0.60

 Yards

-0.57

 revision

-0.57

 Respons

-0.57

FW

-0.56

POSITIVE LOGITS

't

2.00

berra

1.25

adian

1.21

NOT

1.19

 afford

1.18

 easily

1.03

isters

0.96

ny

0.92

 safely

0.90

 manipulate

0.90

Activations Density 0.199%

mentions of the ability to do specific actions or tasks

repetitive use of the word "can."

No Comments

No Known Activations