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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orters

-0.77

ndra

-0.76

akings

-0.75

atri

-0.72

verty

-0.70

 Chamberlain

-0.70

ndum

-0.67

encing

-0.67

onymous

-0.66

 reinstated

-0.64

POSITIVE LOGITS

waves

0.74

gel

0.70

dust

0.67

ebra

0.66

lins

0.65

sat

0.64

feed

0.63

Si

0.62

=]

0.62

ymes

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.