INDEX

Explanations

words related to legal proceedings or documents

oai_token-act-pair · gpt-3.5-turbo

punctuation and various forms of sentence structure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

tarian

-0.77

agonists

-0.72

 superpower

-0.71

 Rodham

-0.69

avorite

-0.67

IFIED

-0.65

ModLoader

-0.64

 Methodist

-0.64

Magic

-0.62

 cloning

-0.62

POSITIVE LOGITS

 Amen

0.94

liga

0.93

 alle

0.92

 Regist

0.83

qui

0.83

ja

0.81

please

0.80

 lang

0.80

si

0.78

nda

0.77

Activations Density 0.159%

words related to legal proceedings or documents

punctuation and various forms of sentence structure

No Comments

No Known Activations