INDEX

Explanations

No Explanations Found

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

IFrame

Link

Negative Logits

ymm

-0.70

atz

-0.69

fal

-0.67

_-

-0.65

cv

-0.64

hed

-0.64

access

-0.63

 weakest

-0.62

uper

-0.62

factor

-0.61

POSITIVE LOGITS

lace

0.77

laws

0.74

 Sands

0.72

theless

0.71

 Mits

0.63

 subcontract

0.63

 sweats

0.62

 Marketplace

0.62

 reprint

0.61

umbnails

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.