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

ococ

-0.80

alde

-0.78

qv

-0.77

kn

-0.74

ergic

-0.71

riks

-0.67

 Strongh

-0.67

ktop

-0.67

Ò

-0.66

endon

-0.66

POSITIVE LOGITS

 litigation

0.69

CLASSIFIED

0.66

 amnesty

0.64

 Verse

0.64

igate

0.62

 Unicorn

0.61

 testament

0.61

 graduation

0.60

 Accord

0.60

 waivers

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.