INDEX

Explanations

phrases indicating a transition or progression of events

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "to."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

don

-0.73

Written

-0.72

puff

-0.66

Input

-0.65

 shouldn

-0.65

rat

-0.64

Week

-0.64

didn

-0.63

DIV

-0.62

FML

-0.60

POSITIVE LOGITS

 encompass

0.98

 downright

0.89

wered

0.83

 adulthood

0.81

pless

0.80

iling

0.79

 infinity

0.76

 ensure

0.76

 accommodate

0.75

asted

0.75

Activations Density 0.077%

phrases indicating a transition or progression of events

occurrences of the word "to."

No Comments

No Known Activations