INDEX

Explanations

references to significant physical locations and institutions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

autions

-0.75

indu

-0.57

ITNESS

-0.55

 guesses

-0.54

lopp

-0.54

 persuasion

-0.53

 extremes

-0.52

lap

-0.52

 smoot

-0.52

 strang

-0.51

POSITIVE LOGITS

 indefinitely

0.86

 amid

0.74

 altogether

0.74

 ASAP

0.73

 afloat

0.68

 permanently

0.65

 assets

0.63

 Â®

0.62

 Holdings

0.62

 premises

0.61

Activations Density 0.317%

references to significant physical locations and institutions

No Comments

No Known Activations