INDEX

Explanations

instances of words related to methods or tools for achieving a specific goal

oai_token-act-pair · gpt-3.5-turbo

connections between concepts and the mechanisms to achieve control or means to an end

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Beir

-0.55

 Events

-0.51

Gov

-0.50

 Costs

-0.48

 assigns

-0.48

 Mate

-0.46

ammers

-0.46

 Regions

-0.46

 Charges

-0.45

 Collections

-0.44

POSITIVE LOGITS

etheless

0.59

lier

0.51

 suspic

0.49

 neigh

0.48

ointed

0.48

natureconservancy

0.48

terness

0.48

figure

0.47

 opin

0.47

cil

0.46

Activations Density 6.366%

instances of words related to methods or tools for achieving a specific goal

connections between concepts and the mechanisms to achieve control or means to an end

No Comments

No Known Activations