INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

UC

-0.77

Football

-0.72

natureconservancy

-0.72

Bat

-0.70

Visual

-0.68

})

-0.68

GMT

-0.67

ãĥŁ

-0.67

Running

-0.67

 mathemat

-0.66

POSITIVE LOGITS

ienne

0.80

ullivan

0.77

arella

0.74

gerald

0.71

 settlements

0.71

 Decay

0.68

riages

0.68

rieve

0.67

rief

0.67

ropy

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.