INDEX

Explanations

phrases indicating the presence of specific positions or locations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 queues

-0.68

chain

-0.65

ratulations

-0.63

comed

-0.62

reviewed

-0.61

matic

-0.61

 commun

-0.61

wallet

-0.60

thodox

-0.59

sett

-0.59

POSITIVE LOGITS

 halftime

1.05

 fullback

1.02

 least

0.97

 shortstop

0.86

 Rookie

0.85

onement

0.83

letico

0.82

 quarterback

0.81

las

0.81

 rookie

0.81

Activations Density 0.060%

phrases indicating the presence of specific positions or locations

No Comments

No Known Activations