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

idia

-0.65

 warehouse

-0.62

port

-0.61

}.

-0.60

Rod

-0.60

bud

-0.59

 proclaim

-0.58

Lil

-0.58

Website

-0.57

 duct

-0.56

POSITIVE LOGITS

ebus

0.86

 Parenthood

0.85

anski

0.75

tti

0.75

imore

0.73

oeuv

0.71

bidden

0.70

 reperto

0.69

hands

0.67

croft

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.