INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tie

-0.75

tested

-0.72

tein

-0.68

alysed

-0.68

retty

-0.67

DEBUG

-0.66

clud

-0.66

Ĥ¬

-0.65

OGR

-0.65

ANY

-0.64

POSITIVE LOGITS

 Archdemon

0.70

 misogyny

0.66

 chloride

0.61

 Kappa

0.60

ipedia

0.59

 resurg

0.59

 wholesale

0.58

orc

0.58

ernel

0.57

isbury

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.