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

 forge

-0.78

ggles

-0.74

ofi

-0.66

 Forge

-0.65

arsity

-0.64

adelphia

-0.64

ãĥķãĤ©

-0.64

forts

-0.63

itions

-0.61

itals

-0.61

POSITIVE LOGITS

 Eater

0.67

avior

0.66

========

0.65

](

0.64

ogenic

0.63

''''

0.62

anasia

0.61

________________

0.61

**

0.60

rend

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.