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

eret

-0.85

fing

-0.77

ormons

-0.68

idates

-0.68

rollers

-0.67

noon

-0.67

izon

-0.64

fin

-0.64

tarian

-0.62

 Pump

-0.62

POSITIVE LOGITS

 bios

0.73

 foremost

0.71

ulhu

0.65

 aval

0.61

DN

0.59

soever

0.59

 recl

0.59

 consensus

0.59

 RELE

0.59

BMC

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.