INDEX

Explanations

phrases containing the prefix "ex-" followed by a word

oai_token-act-pair · gpt-3.5-turbo

references to former relationships or individuals with the prefix "ex-"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

uden

-0.83

Zup

-0.77

 dances

-0.76

 shading

-0.76

oglu

-0.71

 weighted

-0.67

 emoji

-0.65

 vouchers

-0.65

 largeDownload

-0.65

 diagram

-0.62

POSITIVE LOGITS

girlfriend

1.38

husband

1.32

wife

1.20

employ

1.16

military

1.14

Soviet

1.14

members

1.13

member

1.11

president

1.11

squ

1.10

Activations Density 0.027%

phrases containing the prefix "ex-" followed by a word

references to former relationships or individuals with the prefix "ex-"

No Comments

No Known Activations