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

hess

-0.78

lections

-0.76

ername

-0.73

 dens

-0.65

osponsors

-0.63

rou

-0.62

latable

-0.61

 safe

-0.61

 sandy

-0.60

 additive

-0.60

POSITIVE LOGITS

acas

0.74

kok

0.73

isl

0.72

åį

0.71

Interstitial

0.71

oji

0.71

Ku

0.70

Accessory

0.70

Prev

0.69

èĥ

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.