INDEX

Explanations

the word "Me" in sentences

oai_token-act-pair · gpt-3.5-turbo

the token representing the speaker or first-person perspective

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

paralle

-0.69

 totality

-0.67

pole

-0.65

edged

-0.61

rift

-0.60

>>\

-0.59

 overnight

-0.59

ordinate

-0.57

medium

-0.56

 glim

-0.56

POSITIVE LOGITS

asuring

1.41

asured

1.19

cca

1.18

zzo

1.16

asures

1.11

asure

1.11

lda

1.10

aning

1.06

yers

1.03

ghan

1.02

Activations Density 0.030%

the word "Me" in sentences

the token representing the speaker or first-person perspective

No Comments

No Known Activations