INDEX

Explanations

references to significant events or activities, particularly focused on actions taken at a national level

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

OTS

-0.76

UTF

-0.73

Â¯

-0.66

quickShipAvailable

-0.66

ollah

-0.63

utf

-0.63

acters

-0.62

ï¸ı

-0.62

ELF

-0.61

ãĤ§

-0.58

POSITIVE LOGITS

 hiatus

0.99

stage

0.97

 sale

0.91

 stage

0.87

 vacation

0.84

ibaba

0.84

 board

0.83

 tour

0.83

 patrol

0.83

 autop

0.80

Activations Density 0.018%

references to significant events or activities, particularly focused on actions taken at a national level

No Comments

No Known Activations