INDEX

Explanations

historical and factual information

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 âĢº

-0.60

 Mate

-0.58

',

-0.56

 Volunteer

-0.56

 Balk

-0.56

tered

-0.55

 Anon

-0.55

CPR

-0.54

 Toledo

-0.54

POSITIVE LOGITS

Else

0.77

:(

0.74

mosp

0.73

hesda

0.71

yrinth

0.70

jriwal

0.70

roximately

0.69

Quantity

0.69

swers

0.68

APS

0.67

Activations Density 0.093%

historical and factual information

No Comments

No Known Activations