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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hn

-0.66

apa

-0.64

arf

-0.63

breaker

-0.63

 coordinates

-0.58

nia

-0.58

ãĥ£

-0.58

cale

-0.58

 ashore

-0.57

 paralle

-0.57

POSITIVE LOGITS

MIC

0.75

 Krug

0.69

MIC

0.67

rust

0.67

generic

0.66

ukong

0.65

regor

0.65

 temptation

0.62

eatures

0.62

NET

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.