INDEX

Explanations

references to the Supreme Court

oai_token-act-pair · gpt-3.5-turbo

references to the Supreme Court

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

ku

-0.72

eworld

-0.71

 Horus

-0.70

 Typh

-0.68

 Pyro

-0.68

 Saban

-0.65

 Deer

-0.64

jiang

-0.63

raved

-0.61

Fas

-0.59

POSITIVE LOGITS

 justices

1.24

Justice

0.93

rooms

0.87

 Justice

0.85

room

0.84

 rulings

0.82

 Appeals

0.81

 intervened

0.81

OTUS

0.80

Court

0.78

Activations Density 0.026%

references to the Supreme Court

references to the Supreme Court

No Comments

No Known Activations