INDEX

Explanations

phrases related to technical tasks or functions, likely related to programming or software development

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

Not in Any Lists

No Comments

Negative Logits

iates

-0.90

 encounters

-0.71

ievers

-0.68

 interven

-0.67

 experiences

-0.65

iate

-0.64

 corners

-0.64

ankind

-0.63

 Afterwards

-0.62

eor

-0.62

POSITIVE LOGITS

 probably

0.92

 undoubtedly

0.92

 unclear

0.91

 definitely

0.89

 certainly

0.88

 doubtless

0.85

 basically

0.85

 straightforward

0.85

 reminiscent

0.83

 spelled

0.81

Activations Density 0.205%

phrases related to technical tasks or functions, likely related to programming or software development

No Comments

No Known Activations