INDEX

Explanations

mentions of recordkeeping or access to records, especially in a legal or investigative context

oai_token-act-pair · gpt-3.5-turbo

references to records and documentation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bite

-0.80

Pitt

-0.70

 resort

-0.68

issan

-0.66

joy

-0.65

iflower

-0.64

oldemort

-0.63

oji

-0.62

anism

-0.62

 Sadd

-0.62

POSITIVE LOGITS

 records

1.47

 Records

1.20

 record

1.14

 Record

1.00

record

0.96

ords

0.87

 recorder

0.83

 datas

0.79

 tapes

0.79

ographer

0.78

Activations Density 0.012%

mentions of recordkeeping or access to records, especially in a legal or investigative context

references to records and documentation

No Comments

No Known Activations