INDEX

Explanations

phrases starting with the word "Simply" and related actions or instructions

oai_token-act-pair · gpt-3.5-turbo

the word "Simply" and phrases indicating summaries or straightforward explanations

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

 caution

-0.82

 cipher

-0.71

 embargo

-0.69

 evac

-0.65

 corrobor

-0.63

 eventually

-0.63

abal

-0.63

OIL

-0.61

 evacuation

-0.61

 occasional

-0.60

POSITIVE LOGITS

 Simply

2.71

Simply

2.22

 nutshell

1.42

 Simple

1.41

Simple

1.34

 Essentially

1.09

seless

1.09

simple

1.07

 Basically

1.04

 Truly

0.89

Activations Density 0.043%

phrases starting with the word "Simply" and related actions or instructions

the word "Simply" and phrases indicating summaries or straightforward explanations

No Comments

No Known Activations