INDEX

Explanations

the word "iki"

oai_token-act-pair · gpt-3.5-turbo

the character "Miki" 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 @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 predict

-0.64

rhy

-0.61

 increment

-0.60

 exoner

-0.59

 Compton

-0.59

 rapt

-0.58

 Corm

-0.58

 ground

-0.58

 cent

-0.57

 encl

-0.57

POSITIVE LOGITS

iki

4.59

ika

1.68

iko

1.44

iku

1.40

ik

1.32

isha

1.30

ishi

1.26

imi

1.26

uku

1.17

ikan

1.15

Activations Density 0.011%

the word **"iki"**

the character "Miki" in various contexts

No Comments

No Known Activations

the word "iki"