INDEX

Explanations

proper nouns or names that are well-known or recognizable

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate recognition or notoriety associated with people or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ajo

-0.85

plet

-0.78

otion

-0.77

ertation

-0.73

oples

-0.71

otom

-0.70

ermanent

-0.69

erva

-0.69

alach

-0.68

ensation

-0.67

POSITIVE LOGITS

Ð»

0.83

Ô

0.77

ledged

0.77

ÙĨ

0.74

ãĤ¤

0.72

Offline

0.71

lege

0.69

ãĥĥãĥĪ

0.68

abouts

0.67

itarian

0.66

Activations Density 0.032%

proper nouns or names that are well-known or recognizable

phrases that indicate recognition or notoriety associated with people or entities

No Comments

No Known Activations