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

 Icelandic

-0.69

 yield

-0.66

amboo

-0.62

 anarchist

-0.61

 Finnish

-0.61

 Rouhani

-0.60

MIT

-0.60

 occurrence

-0.59

 Italian

-0.59

ID

-0.58

POSITIVE LOGITS

WIND

0.83

ocket

0.80

dp

0.77

ayne

0.76

hawks

0.72

Wide

0.69

cow

0.69

å§«

0.69

CLASSIFIED

0.68

ãĥĦ

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.