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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lev

-0.73

-0.67

aya

-0.67

minus

-0.67

clock

-0.67

 Archdemon

-0.66

aho

-0.63

jad

-0.63

syn

-0.63

POSITIVE LOGITS

 acquisitions

0.79

 histories

0.69

ItemTracker

0.67

vernment

0.67

ufact

0.66

 attrition

0.65

 Tong

0.63

 Frequ

0.62

zsche

0.61

 Burk

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.