INDEX

Explanations

phrases related to tasks or assignments

oai_token-act-pair · gpt-3.5-turbo

references to tasks or responsibilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Geo

-0.82

 Vide

-0.67

atin

-0.62

Rap

-0.61

ental

-0.60

 isot

-0.59

eed

-0.57

ides

-0.57

RR

-0.56

 Metallic

-0.56

POSITIVE LOGITS

 task

1.23

task

1.23

 tasks

1.04

Task

1.00

asks

0.99

 Task

0.93

icult

0.88

masters

0.87

ioned

0.83

nob

0.83

Activations Density 0.011%

phrases related to tasks or assignments

references to tasks or responsibilities

No Comments

No Known Activations