INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

pedia

-0.95

pole

-0.73

 Strateg

-0.68

SPA

-0.66

 contributor

-0.64

INTON

-0.64

 Column

-0.63

UTH

-0.61

Deg

-0.61

 Franch

-0.60

POSITIVE LOGITS

ologies

0.74

nuclear

0.67

tro

0.66

sembly

0.66

ocy

0.66

 guiActiveUn

0.65

breeding

0.65

¶ħ

0.65

perty

0.64

oval

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.