INDEX

Explanations

words related to the concept of apologies or apology-like terms

oai_token-act-pair · gpt-3.5-turbo

the prefix "ap" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Hurricanes

-0.87

 Sands

-0.74

 Sins

-0.74

 Roses

-0.72

 Duke

-0.71

 Apprentice

-0.70

WHERE

-0.70

 Benz

-0.68

 Fargo

-0.68

enegger

-0.68

POSITIVE LOGITS

ollo

1.30

otrop

1.20

ocaly

1.10

osite

1.08

arently

1.07

ocry

1.06

alach

1.03

orters

1.02

onent

1.01

osit

1.01

Activations Density 0.007%

words related to the concept of apologies or apology-like terms

the prefix "ap" in various contexts

No Comments

No Known Activations