INDEX

Explanations

mentions of specific locations or names that include the substring "nom"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "nom" and its variations in relation to nominated entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

gerald

-0.85

EVA

-0.77

 Universe

-0.73

UCT

-0.70

Detailed

-0.69

UID

-0.68

YE

-0.68

 heav

-0.67

GER

-0.66

WI

-0.64

POSITIVE LOGITS

nom

1.38

edy

0.93

asin

0.93

inational

0.92

essage

0.91

onent

0.89

obil

0.88

edia

0.87

undai

0.86

aine

0.86

Activations Density 0.008%

mentions of specific locations or names that include the substring "nom"

occurrences of the word "nom" and its variations in relation to nominated entities

No Comments

No Known Activations