INDEX

Explanations

phrases indicating a particular type of situation or thing without specifying details

oai_token-act-pair · gpt-3.5-turbo

phrases referencing "one of those" scenarios or instances

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

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

obook

-0.80

aeus

-0.78

andise

-0.76

matter

-0.73

kamp

-0.72

analysis

-0.71

querade

-0.70

onis

-0.69

ais

-0.66

culosis

-0.66

POSITIVE LOGITS

 occasions

1.22

 pesky

1.15

 rare

1.14

 times

1.06

 moments

1.05

 situations

1.04

 annoying

1.03

 things

0.96

 instances

0.92

 nights

0.92

Activations Density 0.094%

phrases indicating a particular type of situation or thing without specifying details

phrases referencing "one of those" scenarios or instances

No Comments

No Known Activations