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

resent

-0.70

abiding

-0.69

Cf

-0.67

ĸļ

-0.66

 compr

-0.64

azo

-0.64

avorite

-0.62

accompan

-0.62

ogn

-0.62

issued

-0.59

POSITIVE LOGITS

inity

0.84

LOD

0.79

bots

0.77

vale

0.74

LV

0.74

Planet

0.73

istries

0.72

boy

0.68

mount

0.68

RAW

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.