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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Orig

-0.68

Mah

-0.68

ownt

-0.68

ieth

-0.68

Moe

-0.67

 resc

-0.64

ablishment

-0.64

Yus

-0.61

 reneg

-0.60

olor

-0.59

POSITIVE LOGITS

aptic

0.93

desktop

0.71

shore

0.69

squ

0.68

given

0.67

fitting

0.67

ateurs

0.67

IED

0.65

batch

0.65

gotten

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.