INDEX

Explanations

words related to opening or being opened

oai_token-act-pair · gpt-3.5-turbo

instances of the word "open" and its variations related to physical openings or actions involving doors

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

QL

-0.64

grade

-0.64

ogun

-0.63

Vill

-0.63

gram

-0.62

 Pell

-0.62

 Frequency

-0.62

Score

-0.60

 Depend

-0.60

weet

-0.59

POSITIVE LOGITS

 wounds

0.88

 doors

0.87

 pores

0.85

 holes

0.82

doors

0.79

 curtains

0.77

 windows

0.76

open

0.74

 opened

0.73

 portals

0.73

Activations Density 0.044%

words related to opening or being opened

instances of the word "open" and its variations related to physical openings or actions involving doors

No Comments

No Known Activations