INDEX

Explanations

special characters or symbols used in different languages or encoding formats

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

anwhile

-0.81

 Hitchcock

-0.81

 horizont

-0.79

wagen

-0.69

espie

-0.69

enegger

-0.67

 agre

-0.65

sburgh

-0.64

itude

-0.64

 shots

-0.63

POSITIVE LOGITS

²¾

1.47

¾

1.42

´

1.41

ŃĶ

1.33

ł

1.33

ģ«

1.32

Ł

1.30

©¶æ¥µ

1.27

¹

1.27

§

1.27

Activations Density 0.002%

special characters or symbols used in different languages or encoding formats

No Comments

No Known Activations