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

QB

-0.71

:-

-0.65

OTA

-0.63

 Comput

-0.63

GBT

-0.61

 Telesc

-0.60

 Halifax

-0.58

hetically

-0.56

Â£

-0.56

Db

-0.56

POSITIVE LOGITS

xus

0.79

unks

0.77

wid

0.74

olulu

0.74

eness

0.73

packing

0.73

icho

0.72

 wors

0.69

 Canaver

0.69

Ĥª

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.