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

estones

-0.80

ploma

-0.79

zx

-0.78

oys

-0.75

Track

-0.74

enf

-0.74

inence

-0.73

wcs

-0.72

wer

-0.71

wen

-0.71

POSITIVE LOGITS

 afore

0.63

 catapult

0.60

 rested

0.59

 BOOK

0.59

 elevation

0.59

 benches

0.59

 vegetation

0.58

 Morrow

0.57

 renters

0.57

 predicts

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.