INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 evict

-0.74

ļéĨĴ

-0.64

 paperwork

-0.64

 lease

-0.63

 enlist

-0.63

nas

-0.63

ruce

-0.62

 relocate

-0.61

 shedding

-0.60

 embed

-0.60

POSITIVE LOGITS

aic

0.82

inent

0.80

ãĥ£

0.80

ussen

0.74

[|

0.69

abouts

0.67

||||

0.67

swing

0.66

hover

0.66

Whe

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.