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

è¦ļéĨĴ

-0.76

 FANT

-0.74

MAX

-0.67

CN

-0.65

rons

-0.65

 Ital

-0.61

{:

-0.61

 Radiant

-0.60

AP

-0.60

%:

-0.59

POSITIVE LOGITS

quartered

0.76

estyle

0.74

pun

0.69

uctions

0.68

Pear

0.67

vil

0.67

Sing

0.66

ussion

0.65

hiba

0.65

bent

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.