INDEX

Explanations

terms related to physical heaviness or weight

oai_token-act-pair · gpt-3.5-turbo

references to "heavy" objects or concepts related to significant weight or intensity

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.82

 Machina

-0.80

çīĪ

-0.78

otiation

-0.77

icho

-0.75

ades

-0.73

etts

-0.72

arthed

-0.72

duction

-0.71

isphere

-0.71

POSITIVE LOGITS

weights

1.63

weight

1.21

 hitters

1.15

 hitter

1.14

 duty

1.08

 lifting

1.07

 weights

1.01

metal

0.92

hearted

0.92

 weight

0.92

Activations Density 0.027%

terms related to physical heaviness or weight

references to "heavy" objects or concepts related to significant weight or intensity

No Comments

No Known Activations