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

ibliography

-0.75

pedia

-0.73

---

-0.73

************

-0.70

arts

-0.70

---------

-0.69

Posts

-0.68

adena

-0.67

ricks

-0.67

 Oxford

-0.67

POSITIVE LOGITS

 suits

0.99

 suit

0.98

 relief

0.92

pse

0.71

 async

0.68

£ı

0.66

 rejuven

0.65

HCR

0.64

 waivers

0.63

 serv

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.