INDEX

Explanations

references to multidimensional concepts or objects

oai_token-act-pair · gpt-3.5-turbo

occurrences of the prefix "mult" suggesting multiplicity or variety

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

OPLE

-0.75

 Roundup

-0.69

utenant

-0.68

 pounding

-0.65

 Sorceress

-0.64

 Warden

-0.63

 Villa

-0.63

 mart

-0.62

 Passage

-0.62

 Yankee

-0.62

POSITIVE LOGITS

icol

1.44

ilingual

1.28

itud

1.28

ivariate

1.27

ilateral

1.25

iversal

1.18

itudes

1.17

ivari

1.16

ipl

1.13

imedia

1.13

Activations Density 0.016%

references to multidimensional concepts or objects

occurrences of the prefix "mult" suggesting multiplicity or variety

No Comments

No Known Activations