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

 istg

-0.63

 suicide

-0.62

Pak

-0.62

 bats

-0.61

 Became

-0.60

)]

-0.60

 Dating

-0.59

ulk

-0.58

 illiter

-0.57

UPS

-0.57

POSITIVE LOGITS

ãĥĭ

0.90

ãĤ´ãĥ³

0.81

ĸļ

0.80

DAQ

0.79

cone

0.76

èª

0.72

une

0.72

ONT

0.70

ãĥ´ãĤ¡

0.70

illon

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.