INDEX

Explanations

mentions of specific locations or people involved in legal or criminal proceedings

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

 STATS

-0.66

 quar

-0.65

rency

-0.61

 Laos

-0.58

 recount

-0.58

steen

-0.57

 reprint

-0.57

 nomine

-0.56

issance

-0.55

 veget

-0.55

POSITIVE LOGITS

oshenko

0.87

culosis

0.86

edo

0.81

asury

0.79

orrow

0.76

ulhu

0.74

Mobile

0.70

obl

0.70

nel

0.70

hiba

0.69

Activations Density 1.503%

mentions of specific locations or people involved in legal or criminal proceedings

No Comments

No Known Activations