INDEX

Explanations

mentions of judges in legal contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of judges and legal rulings

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulus

-0.72

 Roads

-0.71

 Rivals

-0.70

haar

-0.68

reek

-0.66

ohyd

-0.65

vised

-0.64

icidal

-0.64

iliar

-0.63

ulated

-0.62

POSITIVE LOGITS

 presiding

0.94

 overseeing

0.94

rooms

0.90

ģĸ

0.86

 examiner

0.82

 Advocate

0.82

 rulings

0.82

 ruled

0.77

 judge

0.76

yers

0.76

Activations Density 0.019%

mentions of judges in legal contexts

mentions of judges and legal rulings

No Comments

No Known Activations