INDEX

Explanations

politicians and political titles

oai_token-act-pair · gpt-3.5-turbo

references to political positions and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 rebound

-0.80

pless

-0.72

anco

-0.71

zb

-0.70

axis

-0.67

ologically

-0.67

sis

-0.67

ãĥ¼ãĥ³

-0.66

ahime

-0.65

 crate

-0.65

POSITIVE LOGITS

 Coun

0.83

 Wendy

0.78

 Gerald

0.77

Coun

0.77

 Joseph

0.76

 Richard

0.76

 Barth

0.75

 Sheila

0.75

Yon

0.74

Vic

0.74

Activations Density 0.088%

politicians and political titles

references to political positions and titles

No Comments

No Known Activations