INDEX

Explanations

references to legal professionals in authoritative roles

oai_token-act-pair · gpt-3.5-turbo

mentions of judges or judicial references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

haar

-0.75

reek

-0.74

ulated

-0.72

ohyd

-0.70

uld

-0.70

uilt

-0.70

kefeller

-0.70

vised

-0.69

irst

-0.67

jri

-0.66

POSITIVE LOGITS

 Advocate

0.92

 examiner

0.89

 presiding

0.87

rooms

0.85

 overseeing

0.82

ģĸ

0.78

 judge

0.78

jud

0.77

 rulings

0.77

 Gorsuch

0.76

Activations Density 0.017%

references to legal professionals in authoritative roles

mentions of judges or judicial references

No Comments

No Known Activations