INDEX

Explanations

acronyms related to health and sciences

oai_token-act-pair · gpt-3.5-turbo

references to a specific organization or project related to "GH"

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

*/(

-0.80

nery

-0.77

andem

-0.74

Parent

-0.74

 parent

-0.70

vier

-0.69

pron

-0.67

sell

-0.67

 Mour

-0.66

pre

-0.64

POSITIVE LOGITS

GH

1.20

GH

1.08

OST

0.97

urst

0.87

arrison

0.84

gh

0.83

OME

0.79

UGE

0.77

APTER

0.77

LLOW

0.75

Activations Density 0.004%

acronyms related to health and sciences

references to a specific organization or project related to "GH"

No Comments

No Known Activations