INDEX

Explanations

phrases describing someone or something being described or considered by others

oai_token-act-pair · gpt-3.5-turbo

phrases indicating attribution or description of subjects by various groups or individuals

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

hent

-0.81

boxing

-0.76

hip

-0.75

shot

-0.75

/"

-0.71

Interstitial

-0.71

lot

-0.70

hea

-0.69

rame

-0.68

hene

-0.68

POSITIVE LOGITS

 critics

1.22

 detractors

1.21

 historians

1.17

 proponents

1.15

 experts

1.12

 scholars

1.11

 commentators

1.10

 economists

1.07

 pundits

1.04

 analysts

1.02

Activations Density 0.154%

phrases describing someone or something being described or considered by others

phrases indicating attribution or description of subjects by various groups or individuals

No Comments

No Known Activations