INDEX

Explanations

references to opposing parties or factions in a conflict

oai_token-act-pair · gpt-4o-mini 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

gee

-0.66

 Rapids

-0.65

 Miko

-0.65

glomer

-0.63

DERR

-0.62

nce

-0.61

Delivery

-0.61

inary

-0.59

 Countdown

-0.58

 Assist

-0.58

POSITIVE LOGITS

 alike

1.07

 thereof

0.97

 sides

0.89

 equally

0.83

 vying

0.82

 concede

0.79

 factions

0.76

terness

0.75

 agree

0.75

 combatants

0.75

Activations Density 0.045%

references to opposing parties or factions in a conflict

No Comments

No Known Activations