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

ãĥ¼ãĥ³

-0.74

 DRAGON

-0.65

 illum

-0.64

 Underground

-0.62

ragon

-0.62

ŃĶ

-0.62

 twitch

-0.61

 nephew

-0.60

 subordinates

-0.60

 cradle

-0.59

POSITIVE LOGITS

gaard

0.82

gettable

0.82

pell

0.80

xus

0.77

 referen

0.75

ascript

0.75

aque

0.74

FFER

0.73

merce

0.73

itia

0.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.