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

oxide

-0.85

oard

-0.83

urat

-0.79

owler

-0.77

utenberg

-0.77

angled

-0.75

ector

-0.75

rocket

-0.74

scope

-0.74

oking

-0.72

POSITIVE LOGITS

 appet

0.64

 Stafford

0.64

 ausp

0.64

 Tribal

0.63

 intest

0.63

 Indo

0.62

 losers

0.62

 childcare

0.61

 brut

0.60

Dom

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.