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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vernment

-0.82

mosp

-0.80

stown

-0.72

©¶æ

-0.70

¬¼

-0.67

onis

-0.66

 unsus

-0.63

zbek

-0.61

 fallout

-0.61

ety

-0.60

POSITIVE LOGITS

 although

0.81

 albeit

0.80

 whereas

0.79

aka

0.76

channelAvailability

0.73

 though

0.72

etc

0.71

SPONSORED

0.70

 depending

0.70

...)

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.