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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ashtra

-0.69

Sport

-0.65

$$$$

-0.65

Evil

-0.64

Corp

-0.64

ukong

-0.63

intendent

-0.63

assic

-0.62

ickey

-0.61

aturday

-0.61

POSITIVE LOGITS

otos

0.71

 Changing

0.71

 briefings

0.63

 experien

0.62

TextColor

0.61

cha

0.61

osit

0.60

Khe

0.60

 sounding

0.60

 docking

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.