INDEX

Explanations

phrases indicating direct speech or quotes

oai_token-act-pair · gpt-3.5-turbo

statements or phrases that contain specific legal or formal phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tremend

-0.74

 assemb

-0.60

 levers

-0.60

 mans

-0.60

 Bronze

-0.58

 dispers

-0.58

 scatter

-0.58

($)

-0.58

 snail

-0.58

nown

-0.57

POSITIVE LOGITS

ľ

1.30

ª

1.22

¬

1.17

¡

1.16

º

1.09

¿

1.08

¦

1.04

Ń

1.04

¤

1.04

Ĵ

1.03

Activations Density 0.198%

phrases indicating direct speech or quotes

statements or phrases that contain specific legal or formal phrases

No Comments

No Known Activations