INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Evidence

-0.70

ournal

-0.68

annis

-0.68

enda

-0.67

gang

-0.67

 Evidence

-0.65

Journal

-0.65

Enlarge

-0.64

 Budd

-0.63

 Jobs

-0.62

POSITIVE LOGITS

renheit

0.89

ãĥĦ

0.74

acle

0.70

 Bengal

0.68

hester

0.67

atar

0.64

ados

0.63

jri

0.63

ãĥ¼ãĤ¯

0.62

 constructor

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.