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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rooms

-0.75

room

-0.70

alyst

-0.69

 Aires

-0.67

etsk

-0.66

 unseen

-0.64

¿½

-0.64

iverpool

-0.64

eva

-0.64

redd

-0.62

POSITIVE LOGITS

VERTISEMENT

0.71

Install

0.70

leneck

0.66

ications

0.63

 Marion

0.63

 Russ

0.61

Joined

0.61

 Marino

0.61

RAFT

0.61

 Claud

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.