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

ogyn

-0.79

abilia

-0.76

ugu

-0.74

elling

-0.74

sted

-0.74

tainment

-0.73

soever

-0.70

antry

-0.68

rican

-0.67

pell

-0.66

POSITIVE LOGITS

veyard

0.82

 suppress

0.64

 supp

0.63

 concent

0.62

activation

0.62

 hypert

0.62

 orbits

0.61

 Mountains

0.61

toggle

0.60

 suppressed

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.