INDEX

Explanations

phrases related to multi-faceted or multi-dimensional entities

oai_token-act-pair · gpt-3.5-turbo

terms related to multi-use or multi-faceted concepts in various contexts

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

Roe

-0.86

 Scouts

-0.75

 kicker

-0.73

rud

-0.69

-0.67

 Ambro

-0.66

 Gors

-0.65

-0.64

 Angels

-0.63

POSITIVE LOGITS

imensional

1.01

asking

0.95

rencies

0.88

acet

0.84

hesive

0.83

format

0.83

edition

0.83

ciplinary

0.83

illion

0.82

ethnic

0.80

Activations Density 0.046%

phrases related to multi-faceted or multi-dimensional entities

terms related to multi-use or multi-faceted concepts in various contexts

No Comments

No Known Activations