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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pload

-0.82

haar

-0.65

 Aval

-0.65

Yad

-0.64

nda

-0.62

 Aram

-0.62

ãĤº

-0.60

ullah

-0.60

oÄŁ

-0.60

 solitary

-0.60

POSITIVE LOGITS

feel

0.73

cheat

0.67

#$

0.66

irtual

0.65

poke

0.63

iculture

0.63

Activ

0.62

Hack

0.61

 reap

0.60

ships

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.