INDEX

Explanations

references to logging data or logging activities

oai_token-act-pair · gpt-3.5-turbo

terms related to logging activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 âĻ

-0.87

á¸

-0.83

 Divine

-0.76

âĻ

-0.71

ollah

-0.69

 Heights

-0.69

ordial

-0.69

 Insp

-0.69

Lev

-0.69

Â½

-0.68

POSITIVE LOGITS

 logging

3.95

 logger

2.05

 logged

1.86

 logs

1.54

 forestry

1.53

log

1.52

 mining

1.33

 timber

1.27

 uploading

1.26

 tagging

1.26

Activations Density 0.012%

references to logging data or logging activities

terms related to logging activity

No Comments

No Known Activations