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

 chars

-0.74

aco

-0.71

ango

-0.66

 slic

-0.65

ensis

-0.63

 caster

-0.63

 projector

-0.62

iatrics

-0.62

angler

-0.61

ict

-0.60

POSITIVE LOGITS

eteenth

0.84

intent

0.81

ãĤ£

0.73

Nap

0.71

Reloaded

0.69

ascript

0.66

Jackson

0.62

inventoryQuantity

0.62

jad

0.61

Eng

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.