INDEX

Explanations

keywords related to subscription services or specific products

oai_token-act-pair · gpt-3.5-turbo

words related to products and services

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Maid

-0.67

 unden

-0.56

 theoret

-0.56

 recip

-0.56

JO

-0.55

 TABLE

-0.55

 streng

-0.53

 priv

-0.53

 proport

-0.51

 explan

-0.50

POSITIVE LOGITS

vre

0.61

KO

0.61

leans

0.59

ierrez

0.57

HF

0.57

Ble

0.57

issance

0.57

entimes

0.57

cliffe

0.57

Morning

0.56

Activations Density 0.013%

keywords related to subscription services or specific products

words related to products and services

No Comments

No Known Activations