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

mi

-0.85

phis

-0.83

STAR

-0.82

stre

-0.76

 isEnabled

-0.76

ross

-0.76

iola

-0.74

PO

-0.73

piece

-0.71

ggies

-0.71

POSITIVE LOGITS

 Helpful

0.65

 Attribution

0.61

VIDIA

0.60

 Rugby

0.59

 CrossRef

0.59

 winters

0.59

 Vote

0.59

geoning

0.59

 Election

0.58

 intu

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.