INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 elig

-0.70

uties

-0.69

WD

-0.68

ounter

-0.67

 PLAN

-0.64

onica

-0.63

asus

-0.61

ivid

-0.61

anamo

-0.60

POSITIVE LOGITS

sung

0.78

vard

0.67

erest

0.65

 arenas

0.60

Gy

0.60

Roh

0.58

visor

0.57

Rub

0.57

 Rosa

0.57

 victory

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.