INDEX

Explanations

references to "place" or the concept of location and positioning

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

 Beans

-0.66

 Lans

-0.63

Suk

-0.59

 Breach

-0.59

MAD

-0.58

cest

-0.58

icia

-0.57

 Briggs

-0.57

edes

-0.57

flix

-0.57

POSITIVE LOGITS

bos

1.10

holders

1.06

holder

0.91

 atop

0.82

 holder

0.77

 amongst

0.77

ngth

0.74

 among

0.72

 beside

0.70

 alongside

0.70

Activations Density 0.015%

references to "place" or the concept of location and positioning

No Comments

No Known Activations