INDEX

Explanations

words related to physical objects used for decoration or as part of a production set

oai_token-act-pair · gpt-3.5-turbo

mentions of "prop" as related to various concepts or contexts in discussions

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

clair

-0.67

hibition

-0.66

 Carbuncle

-0.64

-0.63

 Antar

-0.62

 Clover

-0.62

naire

-0.62

edIn

-0.61

 paths

-0.61

views

-0.60

POSITIVE LOGITS

 tremend

1.04

itious

0.94

etry

0.90

ulsive

0.89

 Gork

0.81

oulos

0.79

undown

0.79

idium

0.79

raints

0.77

 prop

0.75

Activations Density 0.049%

words related to physical objects used for decoration or as part of a production set

mentions of "prop" as related to various concepts or contexts in discussions

No Comments

No Known Activations