INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

SOURCE

-0.75

 propos

-0.73

 indebted

-0.69

 polymorph

-0.68

 Commissioners

-0.67

 heartbeat

-0.65

 contr

-0.62

 envy

-0.61

*)

-0.61

Uri

-0.60

POSITIVE LOGITS

aug

0.88

ousy

0.81

ULTS

0.78

ãĤ¤ãĥĪ

0.77

Ð»

0.76

ãĤ©

0.76

aughter

0.75

byter

0.75

ãĥ¼ãĥĨãĤ£

0.74

kill

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.