INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anova

-0.72

ÙĲ

-0.69

 Rite

-0.66

Err

-0.66

 Benedict

-0.65

Downloadha

-0.63

geist

-0.62

manship

-0.61

 distingu

-0.60

ertodd

-0.60

POSITIVE LOGITS

race

0.84

schild

0.71

apper

0.71

raid

0.71

ipes

0.68

spin

0.68

isexual

0.67

platform

0.67

heat

0.65

skinned

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.