INDEX

Explanations

instances of the word "Tempe"

oai_token-act-pair · gpt-3.5-turbo

mentions of the city or place "Tempe."

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

 Barkley

-0.77

Magikarp

-0.76

 Hodg

-0.73

è£ħ

-0.73

 Jenner

-0.71

 Younger

-0.69

 Scotia

-0.68

atform

-0.65

 Barcl

-0.65

THR

-0.64

POSITIVE LOGITS

poral

1.39

ension

1.14

plates

1.13

porary

1.09

ples

1.07

tem

1.06

perate

1.05

pered

1.02

oshenko

1.01

Tem

0.98

Activations Density 0.009%

instances of the word "Tempe"

mentions of the city or place "Tempe."

No Comments

No Known Activations