INDEX

Explanations

information related to cooking techniques and instructions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

•How

???

No Comments

Negative Logits

idered

-0.62

enture

-0.61

holder

-0.59

ceptions

-0.58

UX

-0.57

Klu

-0.57

Reader

-0.56

entin

-0.54

UA

-0.54

agonists

-0.54

POSITIVE LOGITS

itzer

0.83

--------------------------------------------------------

0.74

beit

0.72

 much

0.71

ling

0.71

ells

0.70

lers

0.67

soever

0.67

 MUCH

0.65

ls

0.65

Activations Density 0.054%

information related to cooking techniques and instructions

No Comments

No Known Activations