INDEX

Explanations

dates in parentheses, possibly related to events or releases

oai_token-act-pair · gpt-3.5-turbo

opening parentheses in the document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 nitrogen

-0.72

 impeachment

-0.70

 spir

-0.69

 electr

-0.69

 contend

-0.68

 nuts

-0.68

 handlers

-0.68

 entitle

-0.68

 pale

-0.68

 vigil

-0.66

POSITIVE LOGITS

?)

1.30

...)

1.28

formerly

1.27

aka

1.24

optional

1.23

via

1.17

includes

1.17

feat

1.16

excluding

1.14

also

1.13

Activations Density 0.150%

dates in parentheses, possibly related to events or releases

opening parentheses in the document

No Comments

No Known Activations