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

zanne

-0.82

bern

-0.79

 thriving

-0.76

 Foss

-0.74

 Tanz

-0.68

steen

-0.67

aepernick

-0.65

wikipedia

-0.65

radical

-0.65

ternity

-0.65

POSITIVE LOGITS

reads

0.80

ips

0.73

 follows

0.70

oon

0.70

ously

0.70

icles

0.68

ence

0.68

ous

0.67

uin

0.65

''.

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.