INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 defects

-0.66

 mutations

-0.65

INTON

-0.62

 critics

-0.62

 dwarf

-0.62

 hamstring

-0.61

 exceptions

-0.61

 reperc

-0.61

 instincts

-0.61

 rival

-0.61

POSITIVE LOGITS

raid

0.76

ository

0.73

uilding

0.73

chens

0.72

otropic

0.72

ãĥĺ

0.70

gard

0.70

fix

0.67

Blu

0.67

Featured

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.