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

 grapes

-0.73

wagen

-0.73

agher

-0.69

ebook

-0.67

 unsur

-0.65

brand

-0.65

MAP

-0.64

selling

-0.63

 gateway

-0.62

search

-0.61

POSITIVE LOGITS

 Holo

0.71

perty

0.71

othy

0.69

ito

0.69

ieth

0.69

anu

0.66

Ic

0.66

iff

0.66

 ILCS

0.65

ythm

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.