INDEX

Explanations

words related to concepts of directionality and causality

oai_token-act-pair · gpt-3.5-turbo

concepts related to various types of "ality," such as morality, legality, and functionality

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

sonian

-0.89

kers

-0.83

edes

-0.79

fman

-0.79

king

-0.75

iors

-0.73

ribut

-0.72

arte

-0.71

href

-0.70

eric

-0.70

POSITIVE LOGITS

istically

0.71

 contag

0.69

 Cloak

0.68

 Rica

0.65

endi

0.61

 Scotia

0.60

butt

0.60

 pact

0.58

 lapse

0.58

hound

0.58

Activations Density 0.040%

words related to concepts of directionality and causality

concepts related to various types of "ality," such as morality, legality, and functionality

No Comments

No Known Activations