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

Not in Any Lists

No Comments

Negative Logits

-0.85

ractor

-0.71

olo

-0.68

ERROR

-0.67

regor

-0.67

erness

-0.65

rob

-0.65

:(

-0.64

ionic

-0.64

aspx

-0.64

POSITIVE LOGITS

 Masquerade

0.76

 horm

0.69

ãĥ´

0.67

 arra

0.66

nces

0.61

 ejac

0.61

 Vers

0.61

nai

0.58

ki

0.58

neapolis

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.