INDEX

Explanations

phrases associated with monetary transactions

oai_token-act-pair · gpt-3.5-turbo

punctuations and conjunctions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

thin

-0.74

dies

-0.68

urus

-0.67

ben

-0.64

DERR

-0.63

fine

-0.59

ribune

-0.59

 warranties

-0.59

fff

-0.58

 Spirits

-0.58

POSITIVE LOGITS

OULD

0.83

hammad

0.81

ItemTracker

0.71

ichick

0.70

é»Ĵ

0.68

 belongs

0.68

ignty

0.67

 Zimmer

0.66

ATURE

0.66

 MUST

0.65

Activations Density 0.571%

phrases associated with monetary transactions

punctuations and conjunctions

No Comments

No Known Activations