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

 Baghd

-0.66

nered

-0.64

ghost

-0.63

spring

-0.63

brand

-0.62

chlor

-0.61

ointed

-0.60

toc

-0.60

store

-0.60

pub

-0.60

POSITIVE LOGITS

 "$:/

0.77

++++++++

0.71

tions

0.70

 Played

0.69

++++++++++++++++

0.65

 Titanic

0.65

 ACTIONS

0.64

à¤

0.63

Continue

0.63

Hep

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.