INDEX

Explanations

phrases related to legal proceedings and actions

oai_token-act-pair · gpt-3.5-turbo

legal terms and concepts related to court cases and 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

 nerds

-0.79

 highs

-0.76

mire

-0.76

 vibe

-0.70

 gigs

-0.70

 Slim

-0.69

ranean

-0.69

rossover

-0.69

 Guys

-0.68

romeda

-0.67

POSITIVE LOGITS

 lawfully

1.36

 Plaintiff

1.33

 unlawfully

1.32

 lawful

1.30

 plaintiff

1.21

 Defendant

1.21

 petitioner

1.17

 appellant

1.17

 Plaint

1.16

 unlawful

1.16

Activations Density 1.638%

phrases related to legal proceedings and actions

legal terms and concepts related to court cases and rulings

No Comments

No Known Activations