INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

iture

-0.76

âĶģ

-0.69

tree

-0.64

 Skinner

-0.61

itton

-0.59

reciation

-0.59

aghan

-0.59

igree

-0.58

caster

-0.58

 Devices

-0.58

POSITIVE LOGITS

0.74

 Galile

0.72

iatus

0.71

 Canaver

0.68

 inflated

0.64

Utt

0.61

 confess

0.61

 resc

0.61

 disg

0.60

Camp

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.