INDEX

Explanations

the word "Basic" repeated in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to "basic" concepts or terms

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sway

-0.77

 witness

-0.71

 blame

-0.68

 swayed

-0.68

 witnesses

-0.67

unm

-0.65

ulo

-0.65

 dispos

-0.65

 scapego

-0.63

 sealing

-0.63

POSITIVE LOGITS

 Basic

3.57

Basic

2.66

basic

2.03

 Basics

1.44

 Intermediate

1.30

 Begin

1.29

 Advanced

1.28

 Generic

1.26

 Simple

1.10

 basic

1.09

Activations Density 0.028%

the word "Basic" repeated in various contexts

references to "basic" concepts or terms

No Comments

No Known Activations