INDEX

Explanations

phrases related to the importance or involvement of something in a particular context

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the role or importance of various entities or factors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.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

 pains

-0.69

 locks

-0.69

 acquaintances

-0.68

 roomm

-0.65

bish

-0.64

killed

-0.64

edIn

-0.61

 APPLIC

-0.59

ancies

-0.59

 entrants

-0.59

POSITIVE LOGITS

ordinate

1.13

 determining

1.11

clus

0.99

 regards

0.98

 shaping

0.97

roads

0.94

 relation

0.93

 conjunction

0.92

 facilitating

0.91

ducers

0.90

Activations Density 0.208%

phrases related to the importance or involvement of something in a particular context

phrases indicating the role or importance of various entities or factors

No Comments

No Known Activations