INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

anga

-0.79

paio

-0.78

iable

-0.74

ratulations

-0.72

aina

-0.70

anie

-0.70

arity

-0.68

istani

-0.68

culosis

-0.68

aji

-0.67

POSITIVE LOGITS

ens

0.84

ãĤ´ãĥ³

0.82

ENS

0.66

 toile

0.65

 Interstellar

0.65

 fundament

0.64

ãĢĤ

0.64

ãĥ¼ãĥ³

0.64

ening

0.63

 hereafter

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.