INDEX

Explanations

organizations or individuals listed in various rankings or lists

oai_token-act-pair · gpt-3.5-turbo

references to rankings or lists and their associated placements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 withd

-0.69

idia

-0.66

orce

-0.65

estic

-0.63

 redes

-0.59

bern

-0.58

nw

-0.58

iour

-0.56

 incurred

-0.56

 countered

-0.55

POSITIVE LOGITS

 Forbes

0.98

 lists

0.97

elist

0.86

 list

0.85

 Lists

0.85

phabet

0.83

 List

0.82

our

0.79

 Countdown

0.77

 waivers

0.75

Activations Density 0.218%

organizations or individuals listed in various rankings or lists

references to rankings or lists and their associated placements

No Comments

No Known Activations