INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fri

-0.76

forest

-0.75

asons

-0.72

rien

-0.67

taboola

-0.66

 cryst

-0.66

pins

-0.65

oug

-0.64

 burden

-0.64

atars

-0.64

POSITIVE LOGITS

 Buchanan

0.78

RN

0.74

 Buckley

0.70

 Mellon

0.69

 TRUMP

0.66

 Commando

0.66

 Eliot

0.66

 Bastard

0.64

 McMaster

0.63

 Dartmouth

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.