INDEX

Explanations

connections between language and various contextual changes or statuses

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

isode

-0.64

xit

-0.61

hens

-0.58

 Appears

-0.58

inburgh

-0.57

 Together

-0.57

yssey

-0.56

iris

-0.55

 âĵĺ

-0.54

xon

-0.54

POSITIVE LOGITS

 anyways

0.71

 anyway

0.66

 collateral

0.64

 already

0.63

 scarce

0.61

 cheap

0.60

 cheaper

0.60

 finite

0.60

 pree

0.60

 cheaply

0.59

Activations Density 0.670%

connections between language and various contextual changes or statuses

No Comments

No Known Activations