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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

early

-0.63

ãĥĸ

-0.63

ãĤ¹ãĥĪ

-0.63

 Fram

-0.62

 Trident

-0.61

éĥ

-0.59

 Brewing

-0.59

Plex

-0.59

 Circuit

-0.58

 ward

-0.58

POSITIVE LOGITS

zl

0.83

undai

0.82

uala

0.80

irez

0.79

ĪĴ

0.77

isphere

0.77

Zup

0.75

yss

0.74

husband

0.74

©¶æ

0.73

Activations Density 0.000%

No Known Activations

This feature has no known activations.