INDEX

Explanations

abbreviations and acronyms, particularly related to organizations or standards

oai_token-act-pair · gpt-3.5-turbo

acronyms and codes

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

 eleph

-1.04

 pione

-0.88

oun

-0.77

 teasp

-0.75

 occas

-0.74

 princ

-0.73

 tremend

-0.70

staking

-0.69

 laun

-0.68

vertisement

-0.65

POSITIVE LOGITS

1.26

1.23

1.21

1.18

Activations Density 0.087%

abbreviations and acronyms, particularly related to organizations or standards

acronyms and codes

No Comments

No Known Activations