INDEX

Explanations

mentions of pardons, amnesties, and related legal terms

oai_token-act-pair · gpt-3.5-turbo

references to pardons and amnesty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Bunker

-0.65

opic

-0.65

Connell

-0.64

ritch

-0.63

Fit

-0.63

lass

-0.62

Engineers

-0.62

Princ

-0.61

MAT

-0.60

str

-0.60

POSITIVE LOGITS

 pard

1.54

 pardon

1.32

 amnesty

0.92

0.85

oning

0.83

cies

0.82

oled

0.78

cess

0.77

ardon

0.76

ige

0.76

Activations Density 0.018%

mentions of pardons, amnesties, and related legal terms

references to pardons and amnesty

No Comments

No Known Activations