INDEX

Explanations

keywords related to sensors and technological functionalities

oai_token-act-pair · gpt-3.5-turbo

references to sensor technologies

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Bald

-0.89

enburg

-0.72

adder

-0.70

 Gazette

-0.70

eve

-0.68

giving

-0.67

burning

-0.66

 Horror

-0.66

ingo

-0.66

ership

-0.64

POSITIVE LOGITS

 sensor

1.10

 proble

1.00

Sensor

0.96

 sensors

0.93

 Sensor

0.93

 enthusi

0.89

 subsystem

0.85

 calibration

0.82

 capacitor

0.82

 shutter

0.80

Activations Density 0.007%

keywords related to sensors and technological functionalities

references to sensor technologies

No Comments

No Known Activations