INDEX

Explanations

phrases related to actions, procedures, or processes

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

agra

-0.65

LH

-0.64

arus

-0.62

atcher

-0.61

algia

-0.59

esta

-0.57

ophy

-0.56

addons

-0.56

raltar

-0.56

 Ging

-0.55

POSITIVE LOGITS

 imaginable

0.91

 whatsoever

0.85

 resembling

0.81

 reminiscent

0.72

 ILCS

0.71

 thereafter

0.68

 throughout

0.67

.''.

0.64

.<

0.64

Activations Density 1.512%

phrases related to actions, procedures, or processes

No Comments

No Known Activations