INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pload

-0.84

emouth

-0.69

OY

-0.66

acters

-0.65

achy

-0.64

çĶŁ

-0.60

dx

-0.60

Hz

-0.59

 Pitch

-0.59

 resolution

-0.58

POSITIVE LOGITS

ascript

0.79

mere

0.78

 dexter

0.75

 millenn

0.73

oeuv

0.72

mitt

0.71

 skelet

0.71

ebook

0.71

amaru

0.71

awa

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.