INDEX

Explanations

acronyms or words related to companies, organizations, or technical terms

oai_token-act-pair · gpt-3.5-turbo

abbreviations or acronyms related to media entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

tained

-0.68

========

-0.68

 Strauss

-0.67

============

-0.65

======

-0.62

juries

-0.61

srf

-0.61

uates

-0.60

 brightest

-0.60

 void

-0.59

POSITIVE LOGITS

gt

1.02

erver

1.01

RP

0.93

ettings

0.91

mith

0.90

pace

0.88

NBC

0.87

emonic

0.87

CL

0.87

aunders

0.84

Activations Density 0.008%

acronyms or words related to companies, organizations, or technical terms

abbreviations or acronyms related to media entities

No Comments

No Known Activations