INDEX

Explanations

locations or places, specifically related to cities or regions

oai_token-act-pair · gpt-3.5-turbo

mentions of locations, particularly cities and counties

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

çĦ

-0.84

selves

-0.80

naires

-0.75

naire

-0.73

jee

-0.73

ãĥ¤

-0.71

onomy

-0.71

payer

-0.70

æĸ

-0.69

Uz

-0.69

POSITIVE LOGITS

 Lauderdale

0.93

 Gators

0.90

paio

0.89

 Keys

0.87

ignt

0.87

ocket

0.83

nton

0.81

 Polk

0.79

 Daytona

0.78

 Hurricanes

0.78

Activations Density 0.015%

locations or places, specifically related to cities or regions

mentions of locations, particularly cities and counties

No Comments

No Known Activations