INDEX

Explanations

references to the term "Ho"

oai_token-act-pair · gpt-3.5-turbo

instances of the term "Ho" followed by a number or abbreviation related to Pokémon

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 totality

-0.76

 mileage

-0.76

CPC

-0.67

 nucleus

-0.67

 pathways

-0.67

 pathway

-0.66

 Defin

-0.66

cape

-0.63

CLASSIFIED

-0.62

 GOODMAN

-0.62

POSITIVE LOGITS

pper

1.33

ppy

1.25

pping

1.20

arding

1.14

oper

1.12

pped

1.11

ppers

1.08

xton

1.07

ppo

1.07

ogle

1.05

Activations Density 0.024%

references to the term "Ho"

instances of the term "Ho" followed by a number or abbreviation related to Pokémon

No Comments

No Known Activations