INDEX

Explanations

words related to agreement or alignment with a statement or sentiment

oai_token-act-pair · gpt-3.5-turbo

instances of agreement or concurrence with statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 jails

-0.69

patch

-0.64

 Gram

-0.64

worm

-0.63

 Tycoon

-0.63

worms

-0.63

 Uriel

-0.61

quer

-0.60

scan

-0.59

MRI

-0.59

POSITIVE LOGITS

 arrang

0.81

 acknow

0.81

 agreeing

0.78

 principle

0.77

awaru

0.76

clus

0.76

 waive

0.75

atever

0.75

enance

0.75

 agre

0.73

Activations Density 0.271%

words related to agreement or alignment with a statement or sentiment

instances of agreement or concurrence with statements

No Comments

No Known Activations