INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Logged

-0.65

-0.59

washer

-0.58

?,

-0.58

 Ended

-0.57

mouth

-0.55

 Carney

-0.53

Registered

-0.51

chio

-0.51

bro

-0.50

POSITIVE LOGITS

idal

0.62

schild

0.61

ideon

0.60

 Tanz

0.57

joice

0.57

rum

0.57

SPONSORED

0.56

urat

0.55

 sidel

0.55

undai

0.54

Activations Density 0.000%

No Known Activations

This feature has no known activations.