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

 proportion

-0.78

 fulfil

-0.75

 fulfillment

-0.72

 prem

-0.71

 redress

-0.70

 apologise

-0.70

 indemn

-0.67

 satisfactory

-0.65

 proport

-0.65

 disarm

-0.65

POSITIVE LOGITS

aga

0.70

Äĩ

0.69

 Corpse

0.67

Sea

0.67

file

0.63

 Travels

0.63

input

0.62

mouth

0.62

ungle

0.61

 Jungle

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.