INDEX

Explanations

phrases related to standing up for beliefs or values

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

EStream

-0.75

javascript

-0.67

EStreamFrame

-0.66

selage

-0.65

ilic

-0.65

erm

-0.65

TION

-0.62

eco

-0.62

iHUD

-0.61

phrine

-0.60

POSITIVE LOGITS

still

1.21

 upright

0.90

 tall

0.88

 vigil

0.85

 steadfast

0.84

 atop

0.78

upe

0.77

alone

0.77

anke

0.77

 defiant

0.76

Activations Density 0.336%

phrases related to standing up for beliefs or values

No Comments

No Known Activations