INDEX

Explanations

words related to creation and development, such as "created" and "developed"

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

appro

-0.66

talk

-0.60

VERTISEMENT

-0.59

tnc

-0.58

aminer

-0.58

 bearer

-0.58

MORE

-0.55

 Mandatory

-0.55

 tide

-0.55

ĪĴ

-0.54

POSITIVE LOGITS

 collabor

1.10

 expressly

0.90

by

0.89

 specifically

0.87

 anew

0.85

 cheaply

0.85

 jointly

0.79

 entirely

0.79

 using

0.77

 principally

0.77

Activations Density 0.136%

words related to creation and development, such as "created" and "developed"

No Comments

No Known Activations