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

unia

-0.67

osphere

-0.65

 Eater

-0.65

inda

-0.60

 residences

-0.60

 Soup

-0.60

ertodd

-0.59

 Route

-0.59

omore

-0.59

perse

-0.58

POSITIVE LOGITS

kB

0.66

DRM

0.66

ãĤ´

0.64

kef

0.64

mingham

0.63

bda

0.61

 skirm

0.61

ãĥĩ

0.61

ierrez

0.61

srf

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.