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

fax

-0.87

ombat

-0.86

sbm

-0.73

 Annotations

-0.69

cia

-0.68

ffee

-0.68

scrib

-0.66

senal

-0.64

SPA

-0.60

MA

-0.59

POSITIVE LOGITS

angled

0.74

sheets

0.72

ulative

0.72

last

0.71

 inver

0.68

idious

0.65

 treating

0.63

Total

0.61

ritic

0.61

 knit

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.