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

FT

-0.78

ttle

-0.71

eps

-0.70

MU

-0.69

NES

-0.68

enge

-0.66

ELD

-0.65

Course

-0.64

 partName

-0.63

isch

-0.63

POSITIVE LOGITS

VIDIA

0.67

 reasoning

0.67

icter

0.65

inventoryQuantity

0.65

ificial

0.64

vernment

0.63

urrency

0.63

 troubles

0.62

 enterprises

0.61

 invalid

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.