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

idth

-0.71

-0.70

ikarp

-0.69

ascus

-0.68

DER

-0.65

matter

-0.64

URA

-0.63

GS

-0.60

rollment

-0.58

 inbox

-0.58

POSITIVE LOGITS

;;;;;;;;;;;;

0.68

owski

0.62

deck

0.62

atures

0.62

Eth

0.61

sov

0.60

ãĤ¦

0.59

 Psycho

0.59

 anch

0.57

 Highlander

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.