INDEX

Explanations

terms related to data aggregation

oai_token-act-pair · gpt-3.5-turbo

terms associated with aggregation and the process of combining data or items

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

stroke

-0.77

vous

-0.72

nton

-0.71

 Misty

-0.70

bler

-0.68

pher

-0.67

foot

-0.66

 Elves

-0.65

warm

-0.65

nee

-0.64

POSITIVE LOGITS

 aggregation

0.99

 aggreg

0.95

 aggregate

0.87

regate

0.80

ulates

0.77

ulate

0.74

ulative

0.72

sum

0.70

 conglomer

0.70

 mund

0.68

Activations Density 0.027%

terms related to data aggregation

terms associated with aggregation and the process of combining data or items

No Comments

No Known Activations