INDEX

Explanations

capital letters followed by digits or other capital letters

oai_token-act-pair · gpt-3.5-turbo

references to brands or businesses

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

 altogether

-0.77

 given

-0.69

 stat

-0.67

 caus

-0.65

 proportion

-0.65

 absent

-0.63

 arbitrarily

-0.62

 entirely

-0.61

theless

-0.60

 privileged

-0.60

POSITIVE LOGITS

uctions

1.25

astery

1.22

apes

1.13

ifles

1.12

istries

1.10

ifts

1.08

ymes

1.08

ixture

1.05

ounces

1.04

ixtape

1.03

Activations Density 0.216%

capital letters followed by digits or other capital letters

references to brands or businesses

No Comments

No Known Activations