INDEX

Explanations

references to different classes or categories, particularly those labeled with numbers

oai_token-act-pair · gpt-3.5-turbo

references to classification labels or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

issance

-0.74

 stump

-0.67

 amplify

-0.67

 push

-0.66

orescence

-0.65

 spor

-0.64

orthy

-0.64

 destro

-0.63

 conceive

-0.62

 usher

-0.62

POSITIVE LOGITS

mates

1.08

 Class

1.08

Class

1.03

ifiers

1.03

 Classes

0.99

Loader

0.97

mate

0.97

ifier

0.96

ifications

0.86

Names

0.86

Activations Density 0.007%

references to different classes or categories, particularly those labeled with numbers

references to classification labels or categories

No Comments

No Known Activations