INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

eton

-0.82

odox

-0.76

isks

-0.75

isky

-0.75

SHIP

-0.72

acular

-0.72

IUM

-0.71

arium

-0.71

unn

-0.69

inav

-0.68

POSITIVE LOGITS

 folds

0.62

 collagen

0.59

 gelatin

0.59

 drawer

0.58

 metab

0.58

 millisec

0.58

genic

0.57

 fused

0.57

 vein

0.55

atel

0.55

Activations Density 0.000%

No Known Activations

This feature has no known activations.