INDEX

Explanations

verbs related to creating, enforcing, or expressing a strong stance or statement

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

CLOSE

-0.62

thia

-0.61

assis

-0.61

Fram

-0.59

phrine

-0.58

Mania

-0.57

pac

-0.56

bour

-0.56

heter

-0.56

 Niet

-0.56

POSITIVE LOGITS

 sure

1.23

hift

1.08

 sense

0.88

 landfall

0.87

 strides

0.82

 excuses

0.82

 headlines

0.80

 adjustments

0.80

URE

0.79

 decisions

0.78

Activations Density 0.458%

verbs related to creating, enforcing, or expressing a strong stance or statement

No Comments

No Known Activations