INDEX

Explanations

activities related to attempting or striving for a goal

oai_token-act-pair · gpt-3.5-turbo

instances of the word "trying."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

dylib

-0.71

lot

-0.69

cised

-0.65

requisite

-0.65

 Origin

-0.64

thus

-0.62

cision

-0.62

oola

-0.60

mary

-0.60

eye

-0.58

POSITIVE LOGITS

 unsuccessfully

0.94

 desperately

0.86

 harder

0.72

rador

0.68

":"/

0.67

 vain

0.66

âĶģ

0.66

ADRA

0.66

 frantically

0.65

 tried

0.65

Activations Density 0.033%

activities related to attempting or striving for a goal

instances of the word "trying."

No Comments

No Known Activations