INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Series

-0.77

 externalToEVAOnly

-0.72

Quantity

-0.69

rentice

-0.69

utable

-0.68

ufact

-0.66

VC

-0.65

esar

-0.65

Interstitial

-0.63

ior

-0.63

POSITIVE LOGITS

ħĭ

0.62

bian

0.61

 breat

0.61

 exploits

0.61

birds

0.61

Poe

0.61

 Passenger

0.61

 riding

0.60

heny

0.60

 sights

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.