INDEX

Explanations

phrases related to superlatives, often indicating the largest, most valuable, or highest in a category or attribute

oai_token-act-pair · gpt-3.5-turbo

superlative phrases indicating significance or ranking in a global context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ADRA

-0.72

odge

-0.69

ModLoader

-0.68

akeru

-0.67

rex

-0.65

izont

-0.64

ience

-0.63

Module

-0.62

aleb

-0.59

ĪĴ

-0.59

POSITIVE LOGITS

 today

0.85

âĦ¢:

0.81

 according

0.76

behind

0.76

*.

0.75

 besides

0.73

 EVER

0.73

 ever

0.72

history

0.72

0.71

Activations Density 0.123%

phrases related to superlatives, often indicating the largest, most valuable, or highest in a category or attribute

superlative phrases indicating significance or ranking in a global context

No Comments

No Known Activations