INDEX

Explanations

terms related to social hierarchies or levels

oai_token-act-pair · gpt-3.5-turbo

references to social class or hierarchical positions within society

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

 Machina

-0.90

tnc

-0.83

vous

-0.77

printf

-0.72

Wikipedia

-0.71

arium

-0.70

 Canaver

-0.69

DVD

-0.68

 Newsweek

-0.67

uality

-0.66

POSITIVE LOGITS

most

0.97

ĺħ

0.82

 upper

0.80

 earners

0.80

 floors

0.79

bridge

0.79

 shaft

0.78

 thigh

0.77

 obser

0.77

 denomination

0.76

Activations Density 0.007%

terms related to social hierarchies or levels

references to social class or hierarchical positions within society

No Comments

No Known Activations