INDEX

Explanations

specific entities and their attributes or associations

oai_token-act-pair · gpt-3.5-turbo

names and attributes related to entertainment content and its creators

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

icable

-0.71

iosyncr

-0.71

 reel

-0.70

 inev

-0.70

voc

-0.70

 tremend

-0.69

 heel

-0.69

 awoken

-0.69

 includ

-0.68

icably

-0.68

POSITIVE LOGITS

 Dealer

0.97

 Guys

0.93

 Moves

0.91

 Matter

0.91

 Hands

0.90

 Places

0.90

 Foods

0.89

 Means

0.88

 Families

0.88

 Definition

0.88

Activations Density 0.188%

specific entities and their attributes or associations

names and attributes related to entertainment content and its creators

No Comments

No Known Activations