INDEX

Explanations

words related to fighting or battling back

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

 Newport

-0.67

viz

-0.65

 realm

-0.61

 Danish

-0.60

 pedigree

-0.59

 Dresden

-0.58

 livest

-0.57

 Moines

-0.57

itarian

-0.56

 Corpse

-0.56

POSITIVE LOGITS

 against

1.01

packs

0.99

lash

0.91

against

0.84

 tears

0.84

GROUND

0.82

stab

0.80

wards

0.76

 forcefully

0.75

assault

0.74

Activations Density 0.027%

words related to fighting or battling back

No Comments

No Known Activations