INDEX

Explanations

the word "noble" with varying degrees of strength

oai_token-act-pair · gpt-3.5-turbo

references to nobility and noble qualities

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

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

DERR

-0.83

è¦ļéĨĴ

-0.66

 Glock

-0.65

 Affect

-0.65

 Virus

-0.64

 Gutenberg

-0.64

CV

-0.63

isp

-0.62

 negatively

-0.62

Accessory

-0.62

POSITIVE LOGITS

 noble

0.97

minded

0.91

cia

0.83

 arist

0.80

 minded

0.77

 pursuits

0.77

 deeds

0.77

 indignation

0.77

bridge

0.76

ments

0.75

Activations Density 0.009%

the word "noble" with varying degrees of strength

references to nobility and noble qualities

No Comments

No Known Activations