INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ersed

-0.77

 Electronics

-0.67

DN

-0.67

 Telecom

-0.66

oried

-0.65

Rx

-0.64

NM

-0.63

 Telecommunications

-0.63

 Attorney

-0.63

 GDDR

-0.62

POSITIVE LOGITS

 lett

0.84

 Ashe

0.76

 apes

0.70

Agent

0.66

 Helena

0.66

 pinch

0.65

 helicop

0.64

 Minerva

0.64

 Xander

0.63

rices

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.