INDEX

Explanations

images and data related activations

oai_token-act-pair · gpt-3.5-turbo

technical terms related to imaging and observational data in astronomy

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

IFrame

Link

Negative Logits

 pacif

-0.72

Style

-0.71

otine

-0.71

Boss

-0.68

cakes

-0.67

tier

-0.66

sac

-0.66

 Styles

-0.66

 Revival

-0.65

 forfeit

-0.65

POSITIVE LOGITS

 telescopes

1.09

 astronomers

1.09

 Telescope

1.06

 telescope

1.04

 datasets

1.00

rared

0.96

 observations

0.93

 imaging

0.93

 Imaging

0.93

 astronomer

0.92

Activations Density 0.191%

images and data related activations

technical terms related to imaging and observational data in astronomy

No Known Activations