INDEX

Explanations

numerical rankings or positions in lists

oai_token-act-pair · gpt-3.5-turbo

ordinal rankings and numerical positions related to various entities

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

bda

-0.70

kamp

-0.65

cember

-0.65

¬¼

-0.64

quished

-0.61

cffff

-0.61

arov

-0.61

instein

-0.60

stals

-0.58

anyahu

-0.57

POSITIVE LOGITS

 largest

1.47

largest

1.45

highest

1.38

 busiest

1.33

 highest

1.31

 longest

1.31

 fastest

1.30

 wealthiest

1.25

 oldest

1.25

 richest

1.25

Activations Density 0.061%

numerical rankings or positions in lists

ordinal rankings and numerical positions related to various entities

No Comments

No Known Activations