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

Not in Any Lists

No Comments

Negative Logits

 Wasserman

-0.74

ATA

-0.65

LIN

-0.65

anguage

-0.63

ABE

-0.61

 conventions

-0.61

orter

-0.59

 blockade

-0.58

byn

-0.58

 periphery

-0.57

POSITIVE LOGITS

aic

0.79

amic

0.77

ingen

0.76

atable

0.74

pire

0.69

ec

0.68

piring

0.67

 Tradable

0.67

bach

0.67

 Norn

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.