INDEX

Explanations

references to "ubs" or "UB" within the text

oai_token-act-pair · gpt-3.5-turbo

references to various types of "subs" or subcategories in a specific context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Evening

-0.63

 Pent

-0.63

 fortune

-0.61

 Directions

-0.60

 successive

-0.59

 ling

-0.59

 Paraly

-0.59

 Murd

-0.58

 Forensic

-0.58

AFP

-0.58

POSITIVE LOGITS

ubs

4.70

ub

2.29

UB

1.93

ubes

1.75

ubby

1.54

ube

1.42

ubi

1.36

ubb

1.33

uber

1.13

unts

1.10

Activations Density 0.013%

references to "ubs" or "UB" within the text

references to various types of "subs" or subcategories in a specific context

No Comments

No Known Activations