INDEX

Explanations

words related to the name "Jac" or variations of it

oai_token-act-pair · gpt-3.5-turbo

references to a character or name related to the story

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

anguage

-0.88

ainment

-0.72

 Skydragon

-0.72

é¾

-0.70

 SPORTS

-0.65

backer

-0.65

skirts

-0.64

origin

-0.64

 gearing

-0.62

IFIED

-0.62

POSITIVE LOGITS

lyn

1.05

Jac

0.99

Jac

0.97

 Jacobs

0.89

obal

0.87

ques

0.84

rium

0.80

quet

0.78

obe

0.78

 Brun

0.75

Activations Density 0.004%

words related to the name "Jac" or variations of it

references to a character or name related to the story

No Comments

No Known Activations