INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

rones

-0.75

--+

-0.69

agi

-0.67

oxide

-0.67

 Tanks

-0.66

 Swordsman

-0.65

gemony

-0.65

enger

-0.65

emis

-0.65

ility

-0.63

POSITIVE LOGITS

 arrang

0.76

stitial

0.71

ocument

0.64

wake

0.64

ksh

0.63

 Registered

0.62

 Scand

0.61

arten

0.59

atching

0.57

ellen

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.