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

azi

-0.68

 blot

-0.64

 confer

-0.62

ADE

-0.60

aternal

-0.60

 Brawl

-0.59

 clearance

-0.58

ahime

-0.58

uggle

-0.58

 Ashe

-0.57

POSITIVE LOGITS

rift

0.89

rss

0.87

leep

0.76

hid

0.75

ictions

0.75

han

0.73

iencies

0.73

irrel

0.73

pron

0.73

yss

0.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.