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

å¤

-0.75

akedown

-0.72

Dispatch

-0.70

Els

-0.69

ħ

-0.69

¼

-0.68

mouth

-0.67

boss

-0.67

Deploy

-0.66

circ

-0.66

POSITIVE LOGITS

iseum

0.67

emale

0.67

enegger

0.67

ischer

0.67

 Variant

0.65

idae

0.62

tal

0.61

 accur

0.60

 composite

0.60

 realised

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.