INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

gery

-0.78

 Nile

-0.70

geries

-0.69

heed

-0.69

DW

-0.68

CJ

-0.64

earch

-0.64

TF

-0.62

 defunct

-0.61

anwhile

-0.61

POSITIVE LOGITS

å¥³

0.83

nie

0.82

orb

0.75

aic

0.73

ALD

0.72

bsp

0.68

yll

0.67

oe

0.67

Lie

0.67

PLAY

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.