INDEX

Explanations

references to film or television screens

oai_token-act-pair · gpt-3.5-turbo

terms related to "screen" or the act of appearing on screen in a cinematic context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ipop

-0.79

ortium

-0.68

doms

-0.68

onomic

-0.66

ton

-0.64

toc

-0.62

 Harriet

-0.62

 horrend

-0.61

 grieving

-0.60

ghan

-0.59

POSITIVE LOGITS

plays

1.33

writers

1.15

writer

1.11

writing

1.01

TVs

0.99

screen

0.94

 screens

0.91

room

0.86

 Screen

0.84

Screen

0.82

Activations Density 0.015%

references to film or television screens

terms related to "screen" or the act of appearing on screen in a cinematic context

No Comments

No Known Activations