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

pload

-0.82

leans

-0.77

tainment

-0.75

atto

-0.72

atican

-0.72

 Submission

-0.67

akedown

-0.63

cules

-0.62

wcsstore

-0.61

edIn

-0.60

POSITIVE LOGITS

ochem

0.79

venture

0.66

aimon

0.65

»Ĵ

0.60

 plaque

0.59

oven

0.59

ocial

0.59

 observed

0.58

 Lynd

0.58

ink

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.