INDEX

Explanations

greeting-related words and phrases

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "greet" and its variations, indicating a focus on greetings and welcoming actions

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

words related to the act of greeting

oai_token-act-pair · gpt-4-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

negie

-0.86

cano

-0.78

proxy

-0.71

hub

-0.67

aq

-0.65

moral

-0.64

cyclop

-0.64

appropri

-0.64

efficient

-0.64

eki

-0.64

POSITIVE LOGITS

 greet

1.28

 greeting

1.14

 greeted

1.04

 greets

1.04

 hello

0.92

 goodbye

0.89

hello

0.85

reetings

0.85

enance

0.83

terson

0.82

Activations Density 0.015%

greeting-related words and phrases

occurrences of the word "greet" and its variations, indicating a focus on greetings and welcoming actions

words related to the act of greeting

No Comments

No Known Activations