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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sucks

-0.75

omach

-0.74

 bothering

-0.74

 dictates

-0.73

 plunge

-0.70

icidal

-0.68

 challeng

-0.68

 afloat

-0.66

 ours

-0.65

 brill

-0.64

POSITIVE LOGITS

por

0.85

cham

0.76

 Mehran

0.71

gee

0.70

gio

0.69

roleum

0.68

enegger

0.67

prus

0.66

 UCHIJ

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.