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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yll

-0.84

Peb

-0.83

STEP

-0.78

INTON

-0.75

 Stain

-0.74

agar

-0.74

çī

-0.71

çĶ°

-0.69

PsyNetMessage

-0.69

ãĥ¼ãĥĨ

-0.68

POSITIVE LOGITS

eatured

0.72

 persistence

0.64

 challenged

0.61

 accumulated

0.60

imeter

0.60

 phased

0.59

 capability

0.59

 assignments

0.59

 reinstated

0.58

":{"

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.