INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ciation

-0.75

ominated

-0.74

imity

-0.73

phia

-0.69

pheus

-0.68

bilt

-0.67

college

-0.67

terday

-0.66

anamo

-0.66

reciation

-0.65

POSITIVE LOGITS

 Jung

0.79

wcsstore

0.70

 throttle

0.66

 psychology

0.66

 CLSID

0.65

 administr

0.63

iHUD

0.61

 frustration

0.61

 Golem

0.61

 predis

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.