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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SEA

-0.73

groups

-0.69

Pod

-0.66

ittees

-0.63

trak

-0.63

 Aires

-0.63

LAT

-0.62

ashington

-0.61

months

-0.61

arters

-0.60

POSITIVE LOGITS

hare

0.79

ister

0.77

bh

0.69

mir

0.69

nih

0.69

igent

0.69

raq

0.67

Ak

0.64

mouth

0.63

arak

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.